工程架构层面优化

.gitignore

@@ -61,6 +61,9 @@ install
 *.png
 *.kml
 
+cache
+StitchLog
+
 
 
 # 保留 3rdparty 整个文件夹内容

CMakeLists.txt

@@ -64,24 +64,24 @@ include_directories(tests)
 
 
 # 测试用例1
-add_executable(stitch tests/main.cpp "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
+add_executable(stitch tests/cpp/main.cpp "tests/cpp/S7215/Arith_zhryp.cpp" "tests/cpp/S7215/Arith_zhryp.h")
 target_link_libraries(stitch ${LIB_STITCH})
 
 # 测试用例2
-add_executable(stitch_DJ "tests/DJ/ProcDJ.cpp" "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
+add_executable(stitch_DJ "tests/cpp/DJ/ProcDJ.cpp" "tests/cpp/S7215/Arith_zhryp.cpp" "tests/cpp/S7215/Arith_zhryp.h")
 target_link_libraries(stitch_DJ  ${LIB_STITCH})
 
 # 测试用例3
-add_executable(stitch_Fea "tests/feaStitchTest.cpp" "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
+add_executable(stitch_Fea "tests/cpp/feaStitchTest.cpp" "tests/cpp/S7215/Arith_zhryp.cpp" "tests/cpp/S7215/Arith_zhryp.h")
 target_link_libraries(stitch_Fea  ${LIB_STITCH})
 
 # 测试用例4
-add_executable(stitch_Genaral "tests/stitch_Genaral.cpp" "tests/utils.cpp" "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
+add_executable(stitch_Genaral "tests/cpp/stitch_Genaral.cpp" "tests/cpp/utils.cpp" "tests/cpp/S7215/Arith_zhryp.cpp" "tests/cpp/S7215/Arith_zhryp.h")
 target_link_libraries(stitch_Genaral  ${LIB_STITCH})
 
 
 #S732
-add_executable(stitch_S732_VL "tests/S732/stitch_S732.cpp" "tests/S732/S732.h" "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
+add_executable(stitch_S732_VL "tests/cpp/S732/stitch_S732.cpp" "tests/cpp/S732/S732.h" "tests/cpp/S7215/Arith_zhryp.cpp" "tests/cpp/S7215/Arith_zhryp.h")
 target_link_libraries(stitch_S732_VL  ${LIB_STITCH} ${OpenCV_LIBS})
 
 
@@ -98,7 +98,7 @@ IF(WIN32)
 
 	set(FFMPEG_LIBRARIES "avcodec" "avformat" "avutil" "swscale" "swresample")
 
-	add_executable(stitch_udp "tests/S732/stitch_udp.cpp" "tests/S732/decodedata.h" "tests/S732/decodedata.cpp" "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
+	add_executable(stitch_udp "tests/cpp/S732/stitch_udp.cpp" "tests/cpp/S732/decodedata.h" "tests/cpp/S732/decodedata.cpp" "tests/cpp/S7215/Arith_zhryp.cpp" "tests/cpp/S7215/Arith_zhryp.h")
 	target_link_directories(stitch_udp PUBLIC ${FFMPEG_LIBS_DIR})
 	target_include_directories(stitch_udp PUBLIC ${FFMPEG_INCLUDE_DIRS})
 	target_link_libraries(stitch_udp ${LIB_STITCH}  ${FFMPEG_LIBRARIES} ${OpenCV_LIBS})
@@ -130,39 +130,71 @@ IF(WIN32)
     set(CMAKE_AUTOMOC ON)
     set(CMAKE_AUTORCC ON)
 
-	add_executable(stitch_S7215ts "tests/S7215/TsDecoder.hpp" "tests/S7215/TsPacker.hpp" "tests/S7215/stitch_S7215.cpp" "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
+	add_executable(stitch_S7215ts "tests/cpp/S7215/TsDecoder.hpp" "tests/cpp/S7215/TsPacker.hpp" "tests/cpp/S7215/stitch_S7215.cpp" "tests/cpp/S7215/Arith_zhryp.cpp" "tests/cpp/S7215/Arith_zhryp.h")
 	target_link_directories(stitch_S7215ts PUBLIC ${FFMPEG_LIBS_DIR})
 	target_include_directories(stitch_S7215ts PUBLIC ${FFMPEG_INCLUDE_DIRS})
 	target_link_libraries(stitch_S7215ts ${LIB_STITCH}  ${FFMPEG_LIBRARIES} ${OpenCV_LIBS}  Qt${QT_VERSION_MAJOR}::Core)
 
 ELSE()
-	# ======== FFMPEG配置 ========
-	set(FFMPEG_INCLUDE_DIRS "/usr/include/x86_64-linux-gnu")
-	set(FFMPEG_LIBS_DIR "/usr/lib/x86_64-linux-gnu")
-	include_directories(${FFMPEG_INCLUDE_DIRS})
-	set(FFMPEG_LIBRARIES "avdevice" "avcodec" "avformat" "avutil" "swscale" "swresample")
-	
-	# == Qt
-	find_package(QT NAMES Qt5 REQUIRED COMPONENTS Widgets )
-    find_package(Qt${QT_VERSION_MAJOR} REQUIRED COMPONENTS Widgets )
-    find_package(Qt5 REQUIRED COMPONENTS PrintSupport)
-    set(TS_FILES QGuideArith_zh_CN.ts)
-
-    set(CMAKE_AUTOUIC ON)
-    set(CMAKE_AUTOMOC ON)
-    set(CMAKE_AUTORCC ON)
-
-	add_executable(stitch_S7215ts "tests/S7215/TsDecoder.hpp" "tests/S7215/TsPacker.hpp" "tests/S7215/stitch_S7215.cpp" "tests/S7215/Arith_zhryp.cpp" "tests/S7215/Arith_zhryp.h")
-	target_link_directories(stitch_S7215ts PUBLIC ${FFMPEG_LIBS_DIR})
-	target_include_directories(stitch_S7215ts PUBLIC ${FFMPEG_INCLUDE_DIRS})
-	target_link_libraries(stitch_S7215ts ${LIB_STITCH}  ${FFMPEG_LIBRARIES} ${OpenCV_LIBS}  Qt${QT_VERSION_MAJOR}::Core)
-
+	# # ======== Qt5 配置 ========
+	# find_package(QT NAMES Qt5 REQUIRED COMPONENTS Core Widgets)
+	# find_package(Qt${QT_VERSION_MAJOR} REQUIRED COMPONENTS Core Widgets)
+	
+	# message(STATUS "QT_VERSION_MAJOR: ${QT_VERSION_MAJOR}")
+	# message(STATUS "Qt5_FOUND: ${Qt5_FOUND}")
+	# message(STATUS "Qt5_VERSION: ${Qt5_VERSION}")
+	
+	# # 启用Qt的MOC、UIC、RCC自动处理
+	# set(CMAKE_AUTOUIC ON)
+	# set(CMAKE_AUTOMOC ON)
+	# set(CMAKE_AUTORCC ON)
+
+	# # ======== FFMPEG配置 ========
+	# # 尝试多个可能的FFmpeg安装路径
+	# set(FFMPEG_POSSIBLE_PATHS
+	# 	"/usr"
+	# )
+
+	# set(FFMPEG_INCLUDE_DIRS ${FFMPEG_DIR}/include)
+	# set(FFMPEG_LIBS_DIR ${FFMPEG_DIR}/lib)
+	# set(FFMPEG_LIBRARIES "avdevice" "avcodec" "avformat" "avutil" "swscale" "swresample")
+	# include_directories(${FFMPEG_INCLUDE_DIRS})
+	# message(STATUS "FFmpeg found at: ${FFMPEG_DIR}")
+
+
+
+	# # ======== 构建 stitch_S7215ts 可执行文件 ========
+	# add_executable(stitch_S7215ts 
+	# 	"tests/cpp/S7215/TsDecoder.hpp" 
+	# 	"tests/cpp/S7215/TsPacker.hpp" 
+	# 	"tests/cpp/S7215/stitch_S7215.cpp" 
+	# 	"tests/cpp/S7215/Arith_zhryp.cpp" 
+	# 	"tests/cpp/S7215/Arith_zhryp.h"
+	# )
+	
+	# # 设置链接目录
+	# if(NOT FFMPEG_LIBS_DIR STREQUAL "")
+	# 	target_link_directories(stitch_S7215ts PUBLIC ${FFMPEG_LIBS_DIR})
+	# endif()
+	
+	# # 设置包含目录
+	# if(FFMPEG_INCLUDE_DIRS)
+	# 	target_include_directories(stitch_S7215ts PUBLIC ${FFMPEG_INCLUDE_DIRS})
+	# endif()
+	
+	# # 链接所有库
+	# target_link_libraries(stitch_S7215ts 
+	# 	${LIB_STITCH}  
+	# 	${FFMPEG_LIBRARIES}
+	# 	${OpenCV_LIBS} 
+	# 	Qt${QT_VERSION_MAJOR}::Core
+	# )
 
 ENDIF()
 
 
-
-
+add_executable(stitch_GeoCorrect "tests/cpp/Test_GeoCorrect.cpp")
+target_link_libraries(stitch_GeoCorrect ${LIB_STITCH})
 
 # 可执行文件输出路径
 set(EXECUTABLE_OUTPUT_PATH ${CMAKE_SOURCE_DIR}/Bin)

stitch/CMakeLists.txt

@@ -98,20 +98,20 @@ add_subdirectory(py/pybind11)
 
 pybind11_add_module(
     UStitcher                 # 模块名
-    py/example.cpp               # 您的 C++ 源文件
-    py/public_binding.cpp
-    # 也可以指定更多源文件...
+    py/API_UnderStitch_binding.cpp               # 主绑定文件
+    py/public_binding.cpp                        # 公共结构绑定
+    # cvbind.h 是头文件，不需要单独编译
 )
 
-target_include_directories(UStitcher PUBLIC src  public_include)
+target_include_directories(UStitcher PUBLIC src  public_include py)
 
 target_link_libraries(UStitcher PUBLIC 
     ${LIB_STITCH}     )
 
-set_target_properties(UStitcher PROPERTIES
-    OUTPUT_NAME "UStitcher"   # 强制输出文件名
-    PREFIX ""             # Windows 默认无前缀，这里显式写更安全
-    SUFFIX ".pyd"         # 禁止生成带 ABI tag 的名字
-)
+# set_target_properties(UStitcher PROPERTIES
+#     OUTPUT_NAME "UStitcher"   # 强制输出文件名
+#     PREFIX ""             # Windows 默认无前缀，这里显式写更安全
+#     SUFFIX ".pyd"         # 禁止生成带 ABI tag 的名字
+# )
 
 

stitch/py/API_UnderStitch_binding.cpp

@@ -0,0 +1,53 @@
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+#include "cvbind.hpp"  // 包含 cv::Mat 类型转换支持
+#include "API_UnderStitch.h"
+#include "StitchStruct.h"
+
+namespace py = pybind11;
+
+// 前向声明
+void bind_public_structures(py::module_ &m);
+
+// 绑定 API_UnderStitch 类
+void bind_API_UnderStitch(py::module_ &m) {
+    py::class_<API_UnderStitch>(m, "API_UnderStitch", "视频帧下视地理拼接")
+        .def("Init", &API_UnderStitch::Init, 
+             py::arg("info"),
+             "初始化拼接")
+        .def("SetOutput", &API_UnderStitch::SetOutput,
+             py::arg("name"), py::arg("outdir"),
+             "设置输出标识和路径")
+        .def("Run", py::overload_cast<cv::Mat, FrameInfo>(&API_UnderStitch::Run),
+             py::arg("img"), py::arg("para"),
+             "运行拼接流程 (cv::Mat版本)")
+        .def("Stop", &API_UnderStitch::Stop,
+             "中止拼接流程")
+        .def("SetConfig", &API_UnderStitch::SetConfig,
+             py::arg("config"),
+             "运行参数配置")
+        .def("OptAndOutCurrPan", &API_UnderStitch::OptAndOutCurrPan,
+             "立即优化并输出当前全景图")
+        .def("ExportPanMat", &API_UnderStitch::ExportPanMat,
+             "获取全景图mat")
+        .def("getHomography", &API_UnderStitch::getHomography,
+             py::arg("info"),
+             "获取单应性矩阵")
+        .def_static("Create", &API_UnderStitch::Create,
+                    py::arg("cachedir") = "./cache",
+                    "创建 API_UnderStitch 实例")
+        .def_static("Destroy", &API_UnderStitch::Destroy,
+                    py::arg("obj"),
+                    "销毁 API_UnderStitch 实例");
+}
+
+PYBIND11_MODULE(UStitcher, m) {
+    m.doc() = "下视拼接算法 Python 绑定";
+
+    // 先绑定公共结构体
+    bind_public_structures(m);
+
+    // 再绑定 API_UnderStitch 类
+    bind_API_UnderStitch(m);
+}
+

stitch/py/UStitcher.pyi

@@ -0,0 +1,158 @@
+from __future__ import annotations
+
+from typing import Any, TypeAlias
+
+from numpy.typing import NDArray
+
+ImageLike: TypeAlias = NDArray[Any]
+
+
+class PointBLH:
+    """地理坐标系（单位：度）."""
+
+    B: float
+    L: float
+    H: float
+
+    def __init__(self) -> None: ...
+
+
+class EulerRPY:
+    """RPY 姿态角（单位：度）."""
+
+    fRoll: float
+    fPitch: float
+    fYaw: float
+
+    def __init__(self) -> None: ...
+
+
+class AirCraftInfo:
+    """载体信息."""
+
+    nPlaneID: int
+    stPos: PointBLH
+    stAtt: EulerRPY
+
+    def __init__(self) -> None: ...
+
+
+class CamInfo:
+    """相机信息."""
+
+    nFocus: int
+    fPixelSize: float
+    unVideoType: int
+    dCamx: float
+    dCamy: float
+    fAglReso: float
+
+    def __init__(self) -> None: ...
+
+
+class ServoInfo:
+    """伺服状态."""
+
+    fServoAz: float
+    fServoPt: float
+    fServoAzSpeed: float
+    fServoPtSpeed: float
+
+    def __init__(self) -> None: ...
+
+
+class FrameInfo:
+    """帧内外方位元素."""
+
+    nFrmID: int
+    craft: AirCraftInfo
+    camInfo: CamInfo
+    servoInfo: ServoInfo
+    nEvHeight: int
+    nWidth: int
+    nHeight: int
+
+    def __init__(self) -> None: ...
+
+
+class UPanInfo:
+    """下视全景图配置."""
+
+    m_pan_width: int
+    m_pan_height: int
+    scale: float
+    map_shiftX: float
+    map_shiftY: float
+
+    def __init__(self) -> None: ...
+
+
+class UPanConfig:
+    """下视拼接参数控制."""
+
+    bUseBA: bool
+    bOutFrameTile: bool
+    bOutGoogleTile: bool
+
+    def __init__(self) -> None: ...
+
+
+class API_UnderStitch:
+    """视频帧下视地理拼接."""
+
+    def Init(self, info: FrameInfo) -> UPanInfo:
+        """初始化拼接，返回全景图配置."""
+        ...
+
+    def SetOutput(self, name: str, outdir: str) -> None:
+        """配置输出标识和目录."""
+        ...
+
+    def Run(self, img: ImageLike, para: FrameInfo) -> int:
+        """运行拼接流程（cv::Mat/numpy.ndarray）."""
+        ...
+
+    def Stop(self) -> None:
+        """中止拼接流程."""
+        ...
+
+    def SetConfig(self, config: UPanConfig) -> None:
+        """更新运行参数."""
+        ...
+
+    def OptAndOutCurrPan(self) -> int:
+        """立即优化并输出当前全景图."""
+        ...
+
+    def ExportPanMat(self) -> ImageLike:
+        """获取当前全景图像."""
+        ...
+
+    def getHomography(self, info: FrameInfo) -> ImageLike:
+        """根据帧信息返回单应性矩阵."""
+        ...
+
+    @staticmethod
+    def Create(cachedir: str = "./cache") -> API_UnderStitch:
+        """创建 API_UnderStitch 实例."""
+        ...
+
+    @staticmethod
+    def Destroy(obj: API_UnderStitch) -> None:
+        """销毁 API_UnderStitch 实例."""
+        ...
+
+
+__all__ = [
+    "API_UnderStitch",
+    "AirCraftInfo",
+    "CamInfo",
+    "EulerRPY",
+    "FrameInfo",
+    "ImageLike",
+    "PointBLH",
+    "ServoInfo",
+    "UPanConfig",
+    "UPanInfo",
+]
+

stitch/py/cvbind.hpp

@@ -0,0 +1,232 @@
+// Created  by ausk<jincsu#126.com> @ 2019.11.23
+// Based on
+// https://github.com/ausk/keras-unet-deploy/tree/master/cpp/libunet/cvbind.h
+
+#pragma once
+
+#include <pybind11/numpy.h>
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+#include <opencv2/core/core.hpp>
+#include <opencv2/highgui/highgui.hpp>
+#include <opencv2/imgproc/imgproc.hpp>
+
+// Convert cv::Point, cv::Rest, cv::Mat
+namespace pybind11
+{
+namespace detail
+{
+//! cv::Point <=> tuple(x,y)
+template <>
+struct type_caster<cv::Point>
+{
+    PYBIND11_TYPE_CASTER(cv::Point, _("tuple_xy"));
+
+    // Convert from Python to C++.
+    // Convert the Python tuple object to C++ cv::Point type, and return false
+    // if the conversion fails.
+    // The second argument indicates whether implicit conversions should be
+    // applied.
+    bool load(handle obj, bool)
+    {
+        // Ensure that the passed parameter is of tuple type
+        if (!pybind11::isinstance<pybind11::tuple>(obj))
+        {
+            std::logic_error("Point(x,y) should be a tuple!");
+            return false;
+        }
+
+        // Extract the tuple object from the handle and ensure its length is 2.
+        pybind11::tuple pt = reinterpret_borrow<pybind11::tuple>(obj);
+        if (pt.size() != 2)
+        {
+            std::logic_error("Point(x,y) tuple should be size of 2");
+            return false;
+        }
+
+        // Convert a tuple of length 2 to cv::Point.
+        value = cv::Point(pt[0].cast<int>(), pt[1].cast<int>());
+        return true;
+    }
+
+    // Convert from C++ to Python.  Convert C++ cv::Mat object to tuple,
+    // parameter 2 and parameter 3 are ignored
+    static handle cast(const cv::Point& pt, return_value_policy, handle)
+    {
+        return pybind11::make_tuple(pt.x, pt.y).release();
+    }
+};
+
+// cv::Rect <=> tuple(x,y,w,h)
+template <>
+struct type_caster<cv::Rect>
+{
+    PYBIND11_TYPE_CASTER(cv::Rect, _("tuple_xywh"));
+
+    bool load(handle obj, bool)
+    {
+        if (!pybind11::isinstance<pybind11::tuple>(obj))
+        {
+            std::logic_error("Rect should be a tuple!");
+            return false;
+        }
+
+        pybind11::tuple rect = reinterpret_borrow<pybind11::tuple>(obj);
+        if (rect.size() != 4)
+        {
+            std::logic_error("Rect (x,y,w,h) tuple should be size of 4");
+            return false;
+        }
+
+        value = cv::Rect(rect[0].cast<int>(),
+                         rect[1].cast<int>(),
+                         rect[2].cast<int>(),
+                         rect[3].cast<int>());
+        return true;
+    }
+
+    static handle cast(const cv::Rect& rect, return_value_policy, handle)
+    {
+        return pybind11::make_tuple(rect.x, rect.y, rect.width, rect.height)
+            .release();
+    }
+};
+
+// Convert between cv::Mat and numpy.ndarray.
+//
+// Python supports a general buffer protocol for data exchange between plugins.
+// Let the type expose a buffer view, this allows direct access to the original
+// internal data, often used in matrix types.
+//
+// Pybind11 provides the pybind11::buffer_info type to map the Python buffer
+// protocol (buffer protocol).
+//
+// struct buffer_info {
+//    void* ptr;                      /* Pointer to buffer */
+//    ssize_t itemsize;               /* Size of one scalar */
+//    std::string format;             /* Python struct-style format descriptor
+//    */ ssize_t ndim;                /* Number of dimensions */
+//    std::vector<ssize_t> shape;     /* Buffer dimensions */
+//    std::vector<ssize_t> strides;   /* Strides (in bytes) for each index */
+//};
+
+template <>
+struct type_caster<cv::Mat>
+{
+public:
+    PYBIND11_TYPE_CASTER(cv::Mat, _("numpy.ndarray"));
+
+    //! 1. cast numpy.ndarray to cv::Mat
+    bool load(handle obj, bool)
+    {
+        array b = reinterpret_borrow<array>(obj);
+        buffer_info info = b.request();
+
+        int nh = 1;
+        int nw = 1;
+        int nc = 1;
+        int ndims = info.ndim;
+        if (ndims == 2)
+        {
+            nh = info.shape[0];
+            nw = info.shape[1];
+        }
+        else if (ndims == 3)
+        {
+            nh = info.shape[0];
+            nw = info.shape[1];
+            nc = info.shape[2];
+        }
+        else
+        {
+            throw std::logic_error("Only support 2d, 2d matrix");
+            return false;
+        }
+
+        int dtype;
+        if (info.format == format_descriptor<unsigned char>::format())
+        {
+            dtype = CV_8UC(nc);
+        }
+        else if (info.format == format_descriptor<int>::format())
+        {
+            dtype = CV_32SC(nc);
+        }
+        else if (info.format == format_descriptor<float>::format())
+        {
+            dtype = CV_32FC(nc);
+        }
+        else if (info.format == format_descriptor<double>::format())
+        {
+            dtype = CV_64FC(nc);
+        }
+        else
+        {
+            throw std::logic_error(
+                "Unsupported type, only support uchar, int32, float, double");
+            return false;
+        }
+        value = cv::Mat(nh, nw, dtype, info.ptr);
+        return true;
+    }
+
+    //! Cast cv::Mat to numpy.ndarray
+    static handle cast(const cv::Mat& mat,
+                       return_value_policy,
+                       handle /*defval*/)
+    {
+        std::string format = format_descriptor<unsigned char>::format();
+        size_t elemsize = sizeof(unsigned char);
+        int nw = mat.cols;
+        int nh = mat.rows;
+        int nc = mat.channels();
+        int depth = mat.depth();
+        int type = mat.type();
+        int dim = (depth == type) ? 2 : 3;
+        if (depth == CV_8U)
+        {
+            format = format_descriptor<unsigned char>::format();
+            elemsize = sizeof(unsigned char);
+        }
+        else if (depth == CV_32S)
+        {
+            format = format_descriptor<int>::format();
+            elemsize = sizeof(int);
+        }
+        else if (depth == CV_32F)
+        {
+            format = format_descriptor<float>::format();
+            elemsize = sizeof(float);
+        }
+        else if (depth == CV_64F)
+        {
+            format = format_descriptor<double>::format();
+            elemsize = sizeof(double);
+        }
+        else
+        {
+            throw std::logic_error(
+                "Unsupport type, only support uchar, int32, float, double");
+        }
+
+        std::vector<size_t> bufferdim;
+        std::vector<size_t> strides;
+        if (dim == 2)
+        {
+            bufferdim = {(size_t)nh, (size_t)nw};
+            strides = {elemsize * (size_t)nw, elemsize};
+        }
+        else if (dim == 3)
+        {
+            bufferdim = {(size_t)nh, (size_t)nw, (size_t)nc};
+            strides = {(size_t)elemsize * nw * nc,
+                       (size_t)elemsize * nc,
+                       (size_t)elemsize};
+        }
+        return array(buffer_info(
+                         mat.data, elemsize, format, dim, bufferdim, strides))
+            .release();
+    }
+};
+}  // namespace detail
+}  // namespace pybind11

stitch/py/example.cpp

@@ -1,118 +0,0 @@
-#include <pybind11/pybind11.h>
-#include <pybind11/stl.h>  // 用于绑定 std::string 等
-#include <pybind11/numpy.h>
-#include <pybind11/stl/filesystem.h>
-#include "StitchStruct.h" // 包含 API_UnderStitch 的定义
-#include "Arith_UnderStitch.h"
-#include "Arith_GeoSolver.h"
-// 可能还需要包含其他依赖项，如 OpenCV 的绑定头文件
-
-namespace py = pybind11;
-
-// 声明外部的辅助绑定函数
-void bind_public_structures(py::module_ &m);
-void bind_api_understitch(py::module_ &m);
-
-
-PYBIND11_MODULE(UStitcher, m) {
-
-    m.doc() = "pybind11 bindings for UnderStitcher.";
-
-    // 绑定通用数据结构
-    bind_public_structures(m);
-
-    // 绑定下视模块接口
-    bind_api_understitch(m);
-
-}
-
-
-
-
-
-void bind_api_understitch(py::module_& m) {
-
-    // UPanConfig
-    py::class_<UPanConfig>(m, "UPanConfig", "下视拼接参数控制")
-        .def(py::init<>())
-        .def_readwrite("bUseBA", &UPanConfig::bUseBA, "开启BA")
-        .def_readwrite("bOutFrameTile", &UPanConfig::bOutFrameTile, "输出单帧正射图")
-        .def_readwrite("bOutGoogleTile", &UPanConfig::bOutGoogleTile, "输出谷歌瓦片");
-
-    // FrameInfo
-    py::class_<FrameInfo>(m, "FrameInfo", "帧内外方位元素")
-        .def(py::init<>())
-        .def_readwrite("nFrmID", &FrameInfo::nFrmID, "帧编号，唯一ID")
-        .def_readwrite("craft", &FrameInfo::craft, "载体信息")
-        .def_readwrite("camInfo", &FrameInfo::camInfo, "相机信息")
-        .def_readwrite("servoInfo", &FrameInfo::servoInfo, "伺服状态")
-        .def_readwrite("nEvHeight", &FrameInfo::nEvHeight, "相对高差")
-        .def_readwrite("nWidth", &FrameInfo::nWidth)
-        .def_readwrite("nHeight", &FrameInfo::nHeight);
-
-    // UPanInfo
-    py::class_<UPanInfo>(m, "UPanInfo", "下视全景图配置")
-        .def(py::init<>())
-        .def_readwrite("m_pan_width", &UPanInfo::m_pan_width)
-        .def_readwrite("m_pan_height", &UPanInfo::m_pan_height)
-        .def_readwrite("scale", &UPanInfo::scale, "比例尺")
-        .def_readwrite("map_shiftX", &UPanInfo::map_shiftX, "平移X")
-        .def_readwrite("map_shiftY", &UPanInfo::map_shiftY, "平移Y");
-    // ----------------------------------------------------
-    // 2. 绑定 API_UnderStitch 类
-    // ----------------------------------------------------
-    py::class_<API_UnderStitch>(m, "API_UnderStitch")
-        // py::init() 是不必要的，因为它是抽象类
-
-
-        // 绑定纯虚函数 (这些函数需要在 C++ 派生类中实现)
-        .def("Init", &API_UnderStitch::Init,
-            py::arg("info"), "初始化拼接")
-
-        .def("SetOutput", &API_UnderStitch::SetOutput,
-            py::arg("name"), py::arg("outdir"), "设置输出标识和路径")
-
-        .def("Run", &API_UnderStitch::Run,
-            py::arg("img"), py::arg("para"), "运行拼接流程")
-
-        .def("Stop", &API_UnderStitch::Stop,
-            "中止拼接流程")
-
-        .def("SetConfig", &API_UnderStitch::SetConfig,
-            py::arg("config"), "运行参数配置")
-
-
-        .def("ExportPanAddr", &API_UnderStitch::ExportPanAddr,
-            py::return_value_policy::reference_internal, // 返回引用，确保 C++ 对象不被析构
-            "获取全景图")
-
-        .def("ExportPanMat", &API_UnderStitch::ExportPanMat,
-            py::return_value_policy::copy, // 通常 cv::Mat 返回时是深拷贝
-            "获取全景图cv::Mat")
-
-        //getHomography
-        .def("getHomography", &API_UnderStitch::getHomography,
-            py::return_value_policy::copy, // 通常 cv::Mat 返回时是深拷贝
-            "获取H映射矩阵:从像方到物方")
-
-        //----------------------------------------------------
-        //3. 绑定静态工厂方法 (Factory Methods)
-        //----------------------------------------------------
-
-        // 绑定 Create 方法
-        .def_static("Create", &API_UnderStitch::Create,
-            py::arg("cachedir") = "./cache",
-            py::return_value_policy::take_ownership, // 返回指针，需要 Python 管理其生命周期
-            "创建 API_UnderStitch 实例")
-
-        // 绑定 Destroy 方法
-        // 注意：这里需要 pybind11::arg::none() 来明确指出该函数没有返回值
-        .def_static("Destroy", &API_UnderStitch::Destroy,
-            py::arg("obj"),
-            "销毁 API_UnderStitch 实例")
-
-        ;
-
-
-}
-

stitch/py/public_binding.cpp

@@ -53,6 +53,37 @@ void bind_public_structures(py::module_ &m) {
         .def_readwrite("fServoPt", &ServoInfo::fServoPt, "当前帧伺服俯仰角")
         .def_readwrite("fServoAzSpeed", &ServoInfo::fServoAzSpeed, "当前帧伺服方位角速度")
         .def_readwrite("fServoPtSpeed", &ServoInfo::fServoPtSpeed, "当前帧伺服俯仰角速度");
+
+    // ----------------------------------------------------
+    // C. FrameInfo 和 UPanInfo
+    // ----------------------------------------------------
+
+    // FrameInfo
+    py::class_<FrameInfo>(m, "FrameInfo", "帧内外方位元素")
+        .def(py::init<>())
+        .def_readwrite("nFrmID", &FrameInfo::nFrmID, "帧编号，唯一ID")
+        .def_readwrite("craft", &FrameInfo::craft, "载体信息 (AirCraftInfo)")
+        .def_readwrite("camInfo", &FrameInfo::camInfo, "相机信息 (CamInfo)")
+        .def_readwrite("servoInfo", &FrameInfo::servoInfo, "伺服状态 (ServoInfo)")
+        .def_readwrite("nEvHeight", &FrameInfo::nEvHeight, "相对高差")
+        .def_readwrite("nWidth", &FrameInfo::nWidth, "图像宽度")
+        .def_readwrite("nHeight", &FrameInfo::nHeight, "图像高度");
+
+    // UPanInfo
+    py::class_<UPanInfo>(m, "UPanInfo", "下视全景图配置")
+        .def(py::init<>())
+        .def_readwrite("m_pan_width", &UPanInfo::m_pan_width, "全景图宽度")
+        .def_readwrite("m_pan_height", &UPanInfo::m_pan_height, "全景图高度")
+        .def_readwrite("scale", &UPanInfo::scale, "比例尺")
+        .def_readwrite("map_shiftX", &UPanInfo::map_shiftX, "平移X")
+        .def_readwrite("map_shiftY", &UPanInfo::map_shiftY, "平移Y");
+
+    // UPanConfig
+    py::class_<UPanConfig>(m, "UPanConfig", "下视拼接参数控制")
+        .def(py::init<>())
+        .def_readwrite("bUseBA", &UPanConfig::bUseBA, "开启BA")
+        .def_readwrite("bOutFrameTile", &UPanConfig::bOutFrameTile, "输出单帧正射图")
+        .def_readwrite("bOutGoogleTile", &UPanConfig::bOutGoogleTile, "输出谷歌瓦片");
 }
 
 

stitch/py/simple.cpp

@@ -1,38 +0,0 @@
-#include <pybind11/pybind11.h>
-#include <string>
-#include <opencv2/opencv.hpp>
-#include "StitchStruct.h" // 包含 API_UnderStitch 的定义
-#include "API_UnderStitch.h"
-using namespace cv;
-
-
-namespace py = pybind11;
-
-// 导出的 C++ 函数
-std::string greet(const std::string &name) {
-    return "Hello, " + name + "! pybind11 module loaded successfully!";
-}
-
-cv::Mat getMat()
-{
-    return cv::Mat::zeros(3,3,CV_8UC1);
-}
-
-cv::Mat getH(FrameInfo info)
-{
-    printf("run get H");
-    auto module = API_UnderStitch::Create();
-    return module->getHomography(info);
-}
-
-
-// PYBIND11 模块入口点
-PYBIND11_MODULE(simple_test, m) {
-    m.doc() = "***************A minimal pybind11 test module.*****************";
-
-    // 绑定 greet 函数
-    m.def("greet", &greet, "A function that returns a greeting string.");
-
-    m.def("getMat", &getMat, "A function that returns a greeting string.");
-    m.def("getH",&getH,"getH");
-}

stitch/src/API_GeoCorrect.h

@@ -0,0 +1,31 @@
+#ifdef _WIN32
+#define STD_STITCH_API __declspec(dllexport)
+#else
+#define STD_STITCH_API  __attribute__ ((visibility("default")))
+#endif
+
+#include "StitchStruct.h"
+
+// 地理校正模块
+class API_GeoCorrect
+{
+public:
+    virtual ~API_GeoCorrect() = default;
+
+    virtual void Correct(cv::Mat img, FrameInfo info) = 0;
+
+    virtual void Init(FrameInfo info) = 0;
+
+
+public:
+    static API_GeoCorrect* Create();
+    static void Destroy(API_GeoCorrect* obj);
+};
+
+
+
+
+
+
+
+

stitch/src/API_UnderStitch.h

@@ -32,6 +32,8 @@ public:
     // 运行拼接流程
     virtual SINT32 Run(GD_VIDEO_FRAME_S img, FrameInfo para) = 0;
 
+    virtual SINT32 Run(cv::Mat img, FrameInfo para) = 0;
+
     // 中止拼接流程
     virtual void Stop() = 0;
 

stitch/src/Arith_BATask.cpp

@@ -18,60 +18,6 @@ using namespace ceres;
 #define STABLE_X3 1920
 #define STABLE_Y3 1080
 
-struct HomographyResidual_1
-{
-
-    HomographyResidual_1(const cv::KeyPoint& keypoint_i, const cv::KeyPoint& keypoint_j, const cv::Mat H1, const cv::Mat H2)
-        : keypoint_i_(keypoint_i), keypoint_j_(keypoint_j),Hi0_(H1), Hj0_(H2)
-    {
-    }
-
-    template <typename T>
-    bool operator()(const T* const h_i, const T* const h_j, T* residual) const
-    {
-        typedef Eigen::Matrix<T, 3, 1> EPoint;
-        typedef Eigen::Matrix<T, 3, 3> EMat;
-
-        EMat Mat_H_i;
-        Mat_H_i << T(h_i[0]), T(h_i[1]), T(h_i[2]),
-            T(h_i[3]), T(h_i[4]), T(h_i[5]),
-            T(h_i[6]), T(h_i[7]), T(1.0);
-
-        EMat Mat_H_j;
-        Mat_H_j << T(h_j[0]), T(h_j[1]), T(h_j[2]),
-            T(h_j[3]), T(h_j[4]), T(h_j[5]),
-            T(h_j[6]), T(h_j[7]), T(1.0);
-
-        EPoint p_i(T(keypoint_i_.pt.x), T(keypoint_i_.pt.y), T(1.0));
-        EPoint p_j(T(keypoint_j_.pt.x), T(keypoint_j_.pt.y), T(1.0));
-
-
-        EPoint warp_i = Mat_H_i * p_i;
-        EPoint img_j = Mat_H_j.inverse() * warp_i;
-        EPoint img_j_1 = img_j / img_j[2];
-
-        // 计算残差向量
-        residual[0] = (img_j_1 - p_j).squaredNorm();
-
-
-
-        return true;
-    }
-
-
-private:
-    const cv::KeyPoint keypoint_i_; // 第 i 帧图像中的特征点
-    const cv::KeyPoint keypoint_j_; // 第 j 帧图像中的特征点
-
-    const cv::Mat Hi0_;
-    const cv::Mat Hj0_;
-
-};
-
-
-
-//// H投影残差：将左图投影到右图，最小化同名点全景图上的投影误差。
-//// 问题：缺少尺度约束，H矩阵优化后失去正交性  优点：形式简单
 struct HomographyResidual
 {
 
@@ -108,34 +54,6 @@ struct HomographyResidual
         residual[0] = (img_j_1 - p_j).squaredNorm();
 
 
-        // 2. 不动点位置约束
-        cv::Point2f pS1 = warpPointWithH(Hi0_, cv::Point2f(STABLE_X1, STABLE_Y1));
-        cv::Point2f pS2 = warpPointWithH(Hi0_, cv::Point2f(STABLE_X2, STABLE_Y2));
-        cv::Point2f pS3 = warpPointWithH(Hi0_, cv::Point2f(STABLE_X3, STABLE_Y3));
-
-        EPoint pS1_E{ (T)pS1.x,(T)pS1.y ,T(1.0) };
-        EPoint pS2_E{ (T)pS2.x,(T)pS2.y,T(1.0) };
-        EPoint pS3_E{ (T)pS3.x,(T)pS3.y ,T(1.0) };
-
-        EPoint p_1{ T(STABLE_X1), T(STABLE_Y1), T(1.0) };
-        EPoint p_2{ T(STABLE_X2), T(STABLE_Y2), T(1.0) };
-        EPoint p_3{ T(STABLE_X3), T(STABLE_Y3), T(1.0) };
-
-
-        EPoint P_r1 = Mat_H_i * p_1;
-        EPoint P_r2 = Mat_H_i * p_2;
-        EPoint P_r3 = Mat_H_i * p_3;
-
-        P_r1 /= P_r1[2];
-        P_r2 /= P_r2[2];
-        P_r3 /= P_r3[2];
-
-
-        // 约束投影位置不变
-        residual[1] = (P_r1 - pS1_E).squaredNorm();
-        residual[2] = (P_r2 - pS2_E).squaredNorm();
-        residual[3] = (P_r3 - pS3_E).squaredNorm();
-
 
         return true;
     }
@@ -322,7 +240,7 @@ void BA_Task::OptFrame(vector<KeyType> frameInd,cv::Mat H_map, std::unordered_ma
     // 创建 Ceres 问题
     ceres::Problem problemH;
     ceres::Problem problemSE3;
-    ceres::Problem problemH2;
+
 
     // 添加残差块
     int nParaCnt = 0;//参数组数
@@ -387,9 +305,9 @@ void BA_Task::OptFrame(vector<KeyType> frameInd,cv::Mat H_map, std::unordered_ma
                 // problemH2.AddResidualBlock(cost_function, nullptr, h_list[i], h_list[j]);
 
                                 ceres::CostFunction* cost_function =
-                    new ceres::AutoDiffCostFunction<HomographyResidual_1, 1, 8, 8>(
-                        new HomographyResidual_1(keypoint_i, keypoint_j, Hi0, Hj0));
-                problemH2.AddResidualBlock(cost_function, nullptr, h_list[i], h_list[j]);
+                    new ceres::AutoDiffCostFunction<HomographyResidual, 1, 8, 8>(
+                        new HomographyResidual(keypoint_i, keypoint_j, Hi0, Hj0));
+                problemH.AddResidualBlock(cost_function, nullptr, h_list[i], h_list[j]);
 #endif
 
 #ifdef OPT_SE3
@@ -421,7 +339,7 @@ void BA_Task::OptFrame(vector<KeyType> frameInd,cv::Mat H_map, std::unordered_ma
 
     // 求解
 #ifdef OPT_H
-    ceres::Solve(options, &problemH2, &summary);
+    ceres::Solve(options, &problemH, &summary);
 #endif
 
 #ifdef OPT_SE3

stitch/src/Arith_GeoCorrect.cpp

@@ -0,0 +1,88 @@
+#include "Arith_GeoCorrect.h"
+#include "Arith_GeoSolver.h"
+#include "Arith_CoordModule.h"
+
+API_GeoCorrect* API_GeoCorrect::Create()
+{
+	return new GeoCorrect();
+}
+
+
+void API_GeoCorrect::Destroy(API_GeoCorrect* obj)
+{
+    delete obj;
+}
+
+
+GeoCorrect::GeoCorrect()
+{
+    _GeoSolver = new GeoSolver();
+}
+
+GeoCorrect::~GeoCorrect()
+{
+}
+
+void GeoCorrect::Correct(cv::Mat src, FrameInfo info)
+{
+    info.servoInfo.fServoPt += 90;
+
+    _GeoSolver->SetOriginPoint(info);
+
+// // GT
+//     //_GeoSolver->SetOriginPoint(info);
+//     cv::Mat R = _GeoSolver->Mat_TransENG2uv(info);
+
+//     cv::Mat R_1 = R.inv();
+
+//     // 计算图像中心点的地面坐标
+//     cv::Point2f centerPtInGeo = warpPointWithH(R_1, cv::Point2f(info.nWidth / 2.0f, info.nHeight / 2.0f));
+
+//     PointXYZ ptCurr = { centerPtInGeo.y, -info.nEvHeight, centerPtInGeo.x};
+//     PointXYZ originCGCSPoint = getXYZFromBLH(info.craft.stPos);
+// 	//PointXYZ diff = getNUEXYZFromCGCSXYZ(ptCurr, originCGCSPoint);
+
+//     PointBLH centerPtInBLH = getBLHFromXYZ(getCGCSXYZFromNUEXYZ(ptCurr, originCGCSPoint));
+//     std::cout << "centerPtInBLH: " << centerPtInBLH.B << ", " << centerPtInBLH.L << ", " << centerPtInBLH.H << std::endl;
+
+// // 
+//     cv::Mat H = _GeoSolver->findHomography(info);
+
+//     cv::Point2f pt_Geo = warpPointWithH(H, cv::Point2f(info.nWidth / 2.0f, info.nHeight / 2.0f));
+
+//     PointXYZ ptNUE = { 0 };
+// 	// 本模块使用的地理系是东(x)-北(y)-地(z),需要转换一下
+// 	ptNUE.X = pt_Geo.y;
+// 	ptNUE.Y = -info.nEvHeight;
+// 	ptNUE.Z = pt_Geo.x;
+// 	PointBLH centerPtInBLH_ = getBLHFromXYZ(getCGCSXYZFromNUEXYZ(ptNUE, _GeoSolver->originCGCSPoint));
+//     std::cout << "centerPtInBLH_: " << centerPtInBLH_.B << ", " << centerPtInBLH_.L << ", " << centerPtInBLH_.H << std::endl;
+
+    cv::Mat H = _GeoSolver->findHomography(info);
+
+    cv::Rect2f roi = warpRectWithH_2Rect(H, src.size());
+
+    float scale = src.cols / MAX(roi.width, roi.height);
+    float shiftX = -roi.x * scale;
+    float shiftY = (roi.y  + roi.height)* scale;
+
+    Mat H_proj = (Mat_<double>(3, 3) << scale, 0, shiftX,
+        0, -scale, shiftY,
+        0, 0, 1
+        );
+    cv::Rect2f roi2 = warpRectWithH_2Rect(H_proj * H, src.size());
+
+    cv::Mat dst;
+    warpPerspective(src, dst, H_proj * H, cv::Size(roi2.width, roi2.height));
+
+
+    // 计算图像四至的经纬高
+
+
+    return; 
+}
+
+void GeoCorrect::Init(FrameInfo info)
+{
+    _GeoSolver->SetOriginPoint(info);
+}

stitch/src/Arith_GeoCorrect.h

@@ -0,0 +1,21 @@
+#pragma once
+#include "API_GeoCorrect.h"
+#include "opencv2/opencv.hpp"
+#include "Arith_GeoSolver.h"
+#include "StitchStruct.h"
+#include "FileCache.h"
+#include <map>
+#include "Logger.h"
+
+class GeoCorrect : public API_GeoCorrect
+{
+public:
+    GeoCorrect();
+    ~GeoCorrect();
+
+    void Correct(cv::Mat img, FrameInfo info);
+    void Init(FrameInfo info);
+
+private:
+    GeoSolver* _GeoSolver;
+};

stitch/src/Arith_GeoSolver.cpp

@@ -139,20 +139,12 @@ Mat GeoSolver::Mat_TransENG2uv(FrameInfo info)
 		0, 0, info.nEvHeight
 		);
 
-
-	// 内参
-	FLOAT32 fd = info.camInfo.nFocus / info.camInfo.fPixelSize * 1000;
-
-	Mat M_cam = (Mat_<double>(3, 3) << fd, 0, info.nWidth / 2,
-		0, -fd, info.nHeight / 2,
-		0, 0, 1
-		);
-
-	Mat M = M_cam * Mat_TransENG2Cam(info) * M_het;
+	Mat M = Mat_GetCamK(info) * Mat_TransENG2Cam(info) * M_het;
 
 	return M;
 }
 
+using namespace std;
 cv::Mat GeoSolver::Mat_TransENG2Cam(FrameInfo info)
 {
 

stitch/src/Arith_GeoSolver.h

@@ -62,7 +62,7 @@ public:
 	// 经纬度转换为局部地理系坐标
     cv::Point2f getGeoFromBLH(PointBLH ptPos);
 
-private:
+public:
 
     // 计算当前帧李群（SE3）和投影K
     RtK AnlayseRtK(FrameInfo info);
@@ -76,7 +76,7 @@ private:
     // 机体ENG(东北地)到像方的 旋转矩阵
     cv::Mat Mat_TransENG2uv(FrameInfo info);
 
-private:
+public:
 
     // 平移矩阵，以初始化点为基准，计算当前位置在初始点的地理坐标，那么当前帧所有点的坐标做此平移
     cv::Mat Mat_TransENGMove(FrameInfo info);

stitch/src/Arith_UnderStitch.cpp

@@ -132,7 +132,7 @@ UPanInfo UnderStitch::InitMap(FrameInfo para)
 	auto geo_pro_rect = warpRectWithH_2Rect(H0,cv::Size(para.nWidth, para.nHeight));
 	
 	// 计算目标比例（单帧投影后占全景图的0.25）
-	float target_ratio = 0.25;
+	float target_ratio = 0.08;
 	float current_ratio = MAX(geo_pro_rect.width / (target_ratio*panPara.m_pan_width), geo_pro_rect.height / (target_ratio*panPara.m_pan_height));
 
 	// 调整scale参数
@@ -468,6 +468,49 @@ cv::Mat gammaCorrect(const cv::Mat& src, double gamma)
 }
 
 
+GD_VIDEO_FRAME_S mat_to_gd_frame(const cv::Mat& mat) 
+{
+    GD_VIDEO_FRAME_S frame = {0};
+    
+    if (mat.empty()) {
+        return frame;
+    }
+    
+    frame.u32Width = mat.cols;
+    frame.u32Height = mat.rows;
+    
+    // 根据 Mat 的类型和通道数确定像素格式
+    int channels = mat.channels();
+    int depth = mat.depth();
+    
+    if (channels == 1 && depth == CV_8U) {
+        frame.enPixelFormat = GD_PIXEL_FORMAT_GRAY_Y8;
+    } else if (channels == 3 && depth == CV_8U) {
+        // 假设是 BGR（OpenCV 默认）
+        frame.enPixelFormat = GD_PIXEL_FORMAT_BGR_PACKED;
+    } else if (channels == 4 && depth == CV_8U) {
+        // BGRA
+        frame.enPixelFormat = GD_PIXEL_FORMAT_BGR_PACKED;  // 使用 BGR，忽略 alpha
+    } else {
+        throw std::runtime_error("Unsupported cv::Mat format for GD_VIDEO_FRAME_S");
+    }
+    
+    // 设置 stride 和虚拟地址
+    frame.u32Stride[0] = static_cast<unsigned int>(mat.step[0]);
+    frame.u64VirAddr[0] = mat.data;
+    
+    return frame;
+}
+
+
+SINT32 UnderStitch::Run(cv::Mat img, FrameInfo para)
+{
+	// 将cv::Mat转换为GD_VIDEO_FRAME_S
+	GD_VIDEO_FRAME_S frame = mat_to_gd_frame(img);
+	return Run(frame, para);
+}
+
+
 SINT32 UnderStitch::Run(GD_VIDEO_FRAME_S frame, FrameInfo para)
 {
 	stopWatch sw;

stitch/src/Arith_UnderStitch.h

@@ -39,6 +39,8 @@ public:
 
     SINT32 Run(GD_VIDEO_FRAME_S img, FrameInfo para);
 
+    SINT32 Run(cv::Mat img, FrameInfo para);
+
     void Stop();
 
     SINT32 OptAndOutCurrPan();

stitch/src/Version.h

@@ -2,5 +2,5 @@
 
 #pragma once
 #include <string>
-std::string BUILD_TIME = "BUILD_TIME 2025_11_27-19.34.59";
+std::string BUILD_TIME = "BUILD_TIME 2025_11_29-17.52.06";
 std::string VERSION = "BUILD_VERSION 1.0.1";

tests/1.py

@@ -1,56 +0,0 @@
-import os
-import sys
-
-# 1. 定义您的 DLL 目录
-DLL_DIR = r"D:\wangchongwu_gitea_2023\StitchVideo\Bin"
-CV_DIR = r"C:\Lib\opencv455\build\x64\vc14\bin"
-CUDA_DIR = r"C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.4\bin"
-
-# path
-os.add_dll_directory(DLL_DIR)
-os.add_dll_directory(CV_DIR)
-os.add_dll_directory(CUDA_DIR)
-
-sys.path.append(DLL_DIR)
-sys.path.append(CV_DIR)
-sys.path.append(CUDA_DIR)
-
-# 4. 导入
-from UStitcher import API_UnderStitch,FrameInfo
-
-input("sssss")
-
-
-frame_info = FrameInfo()
-
-
-
-frame_info.nFrmID = 1
-frame_info.craft.stPos.B = 39
-frame_info.craft.stPos.L = 120
-frame_info.craft.stPos.H = 1000
-frame_info.camInfo.nFocus = 40
-frame_info.camInfo.fPixelSize = 12
-
-frame_info.servoInfo.fServoAz = 90
-frame_info.servoInfo.fServoPt = -45
-
-frame_info.nEvHeight = 1200
-
-frame_info.nWidth = 1280
-frame_info.nHeight = 1024
-
-
-
-stitcher = API_UnderStitch.Create()
-#H = stitcher.getHomography(frame_info)
-
-
-print("done")
-
-
-
-
-
-
-

tests/cpp/DJ/ProcDJ.cpp

@@ -371,7 +371,7 @@ int main()
     vector<std::string> videoPathList;
     vector<std::string> srtPathList;
 
-    string folder = "F:/K2D_data/";
+    string folder = "/media/wang/data/K2D_data/";
 
 
     videoPathList.push_back(folder + "DJI_20251024144932_0001_Z.MP4");

tests/cpp/S732/NeoArithStandardDll.h

@@ -0,0 +1,655 @@
+#pragma once
+/*********版权所有(C)2024, 武汉高德红外股份有限公司***************
+* 文件名称：ArithStandardDll.h
+* 文件标识：高德光电搜索跟踪算法SDK
+* 内容摘要：
+* 其它说明：算法动态链接库(Arith DLL)的函数、全局变量、宏定义，统一前缀为简写"ARIDLL"
+* 当前版本：V2.0
+* 创建作者：04046wcw
+* 创建日期：2023-11-01
+****************************************************************/
+#ifndef __NEO_ARTTHSTANDARDDLL_H__
+#define __NEO_ARTTHSTANDARDDLL_H__
+
+#include "PlatformDefine.h"
+#include "Arith_CommonDef.h"
+
+
+
+#ifdef _WIN32
+#define STD_TRACKER_API extern "C" __declspec(dllexport)
+#else
+#define STD_TRACKER_API  __attribute__ ((visibility("default")))
+#endif
+
+
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+
+
+//+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+//单个目标的结构体[兼容输入/输出,检测/跟踪]
+typedef struct tagARIDLL_OBJINFO
+{
+    //*****1.目标状态信息*****
+    int            nFrameId;       //目标当前信息所对应的帧编号
+    unsigned char  unObjStatus;    //目标搜索状态信息，更新/新增/删除
+    unsigned char  bMainTracked;   //目标是否为主跟踪目标
+    TrackingStatus unTrackingStatus;//目标跟踪状态
+
+
+    //*****2.目标管道信息*****
+    int            nOutputID;      //输出告警目标
+    int            nInPipesID;     //目标在管道数组中的编号
+    int            nPipeLostCnt;   //目标当前管道连续丢失计数
+    int            nTotalCnt;      //目标当前管道总帧数
+    unsigned char  bInCurrFov;     //目标是否在当前视场
+    int            nAntiJamming;   //抗干扰状态
+
+    //*****3.目标核心信息*****
+    float          nX;             //目标中心点图像坐标x
+    float          nY;             //目标中心点图像坐标y
+    float          nObjW;          //目标宽度
+    float          nObjH;          //目标高度
+    float          fAz;            //目标当前方位角 
+    float          fPt;            //目标当前俯仰角
+
+    //  目标预测位置
+    float          fPredAz;
+    float          fPredPt;
+
+    //*****4.其他属性信息*****
+    int            nObjGray;       //目标灰度
+    int            nObjMaxGray;    //目标极值灰度
+    int            nMaxPosX;       //目标极大值点X
+    int            nMaxPosY;       //目标极大值点Y
+    int            nPixCnts;       //目标像素个数
+    unsigned char  ubSizeType;     //目标尺寸类型:    
+    float          fProb;          //目标识别置信度
+    float          fSNR;           //目标信噪比值
+    float          fTgEntropy;     //目标信息熵值
+    float          fBgEntropy;     //目标背景信息熵
+    float          fSaliency;      //目标显著性值
+
+    // 
+    bool           nJammingSucess; //目标成功干扰
+    
+    int            unClsType;       //目标类别
+    float          fReIDSim;        //当前目标与主目标的ReID相似度
+
+    // 如果处于跟踪状态，则输出下列值
+    RECT32S         SA_SrBox;       //小面目标跟踪波门
+    SizeType        SA_SizeType;    //尺度信息
+    RECT32S         KCF_SrBox;      //KCF波门
+    RECT32S         TLD_SrBox;      //TLD波门
+    FLOAT32         fConf;          //跟踪置信度
+    ObjSrc          ArithSrc;       //跟踪算法来源，决策后
+
+    unsigned char   byte[20];       //预留
+
+}ARIDLL_OBJINFO;
+
+
+//输入【系统参数】结构体
+typedef struct tagARIDLL_INPUTPARA
+{
+    int             nTimeStamp;                 //当前帧采集时刻时间戳，单位毫秒
+    int             unFrmId;                    //当前帧图像帧编号
+    short           unFreq;                     //输入图像帧频
+    ServoInfo       stServoInfo;                //传感器伺服信息
+    CamInfo         stCameraInfo;               //相机信息
+    AirCraftInfo    stAirCraftInfo;             //载体信息
+    GuideInfo       stGuideInfo;                //外部引导信息
+
+    AIT_OUTPUT      stAITrackerInfo;            //AI跟踪器结果
+
+    int             nServoDelatCnt;             //伺服角度延迟帧数
+    // 其他输入
+    bool            bImageRataSys;              //像旋系统标记，S731实物样机1，数字样机未模拟像旋 -0
+    int             nElevationDiff;             //机载设备挂飞高程差
+}ARIDLL_INPUTPARA;
+
+//调试信息
+typedef struct tagARIDLL_DEBUG_OUTPUT
+{
+    unsigned short  nDetectObjsNum;
+    // 管道资源
+    unsigned short  nMaxPipeNum;     //当前系统管道资源数量
+    unsigned short  nUsePipeNum;     //当前非空管道数量
+    float           Arith_time;      //算法运行耗时
+    unsigned int    unFrmID;         //算法执行的帧编号
+
+    //ARM发指令信息
+    unsigned char  nSysMode;        //外部系统状态
+    unsigned char  nScenMode;       //场景模式
+    unsigned char  nStatus;         //待命/检测/跟踪/丢失状态信息等
+    unsigned char  nPixelType;      //图像数据类型
+    unsigned short nWidth;          //图像宽
+    unsigned short nHeight;         //图像高
+    unsigned char  nLockType;       //1-拉框吸附 2-点选吸附 3-ID锁定 4-修正攻击点 5-解锁
+    unsigned char  nLockID;         //锁定id号
+    unsigned short nLockX;          //锁定波门或者修改攻击点
+    unsigned short nLockY;
+    unsigned short nLockW;
+    unsigned short nLockH;
+    unsigned short nPredictX;       //轨迹预测点
+    unsigned short nPredictY;
+    unsigned short nForceMemFrm;    //强制记忆帧数
+    unsigned char  unFreq;          //帧频
+    float          fServoAz;        //伺服方位角
+    float          fServoPt;        //伺服俯仰角
+    float          nFocus;          //焦距
+    float          fPixelSize;      //像元尺寸
+    unsigned char  unVideoType;     //视频类型
+
+    //算法参数(公用)
+    unsigned short      nX;             //决策输出中心点X
+    unsigned short      nY;             //决策输出中心点Y
+    unsigned short      nW;             //决策输出宽度
+    unsigned short      nH;             //决策输出高度
+
+    unsigned short      nRecapX;        //重捕区域中心X
+    unsigned short      nRecapY;        //重捕区域中心Y
+    unsigned short      nRecapW;        //重捕区域宽度
+    unsigned short      nRecapH;        //重捕区域高度
+
+    //对地跟踪信息调试
+    //对地参数
+    unsigned char       nDecisionStatus;        //决策状态
+    unsigned char       nKcfStatus;             //kcf状态
+    unsigned char       bAIDStatus;             //AI识别状态
+    float                      fKCFRes;                //KCF响应
+    float                     fLargeResTH;            //KCF重捕阈值
+    float                     fArrKCFRes;
+    unsigned char       nOccKCFStatus;
+    unsigned char       nArrestKCFStatus;
+    short               nAIDBestId;
+    short               nAIDLostCnt;
+    short               unContiTrackedCnt;
+    short               nAIDJamCnt;
+    unsigned char       nOccAIDStatus;
+    unsigned char       nArrestAIDStatus;
+    unsigned short      nTLDNum;                //TLD聚类检测个数
+    unsigned short      nLearnCnt;              //TLD学习计数
+    float                      fMaxNNConf;             //TLD检测最大响应
+
+    //对空参数
+    unsigned char       sky_bComplexEnv;        // 复杂背景标志位
+    unsigned char       sky_bInterferenceMem;   // 干扰近记忆标志位
+    unsigned char       sky_bResetByAIDet;      // 跟踪被AI重置标记
+    unsigned char       sky_nClassSource;       // 目标类别输出来源
+    unsigned char       sky_nDecisionStatus;    // 决策状态，输出来源
+    unsigned char       sky_TrkDownRatio;       // 对空跟踪降采样倍数
+    unsigned short      sky_TrkMemFrm;          // 跟踪目标进记忆帧数
+    unsigned short      sky_nTrkCX;             // 决策目标信息中心点X
+    unsigned short      sky_nTrkCY;             // 决策目标信息中心点Y
+    unsigned short      sky_nTrkW;              // 决策目标信息宽度
+    unsigned short      sky_nTrkH;              // 决策目标信息高度
+    unsigned short      sky_nTrkPxlsCnt;        // 跟踪目标像素数
+    unsigned short      sky_fTrkConf;           // 跟踪目标置信度
+    unsigned char       sky_bGuideInFov;        // 导引目标是否在视场
+    unsigned short      sky_nGuideCX;           // 导引区域中心X
+    unsigned short      sky_nGuideCY;           // 导引区域中心Y
+    unsigned short      sky_nGuideW;            // 导引区域宽度
+    unsigned short      sky_nGuideH;            // 导引区域高度
+
+    unsigned char       resv[48];               //预留
+}ARIDLL_DEBUG_OUTPUT;
+
+
+
+//跟踪目标输出结构体
+typedef struct tagARIDLL_OUTPUT
+{
+    int nTimeStamp;//当前帧时间戳(透传)，单位：毫秒
+
+    // 系统工作模式(透传)//           by wcw04046 @ 2021/12/06
+    GLB_SYS_MODE nSysMode;
+    int nFrmNum;//处理帧计数
+
+    // 场景模式
+    GLB_SCEN_MODE nScenMode;
+
+    //*****工作状态*****
+    GLB_STATUS  nStatus;                                //待命/检测/跟踪/丢失状态信息等
+
+    //*****目标检测*****(短时航迹点，用于用户指示)
+    int nAlarmObjCnts;                                  //当前帧告警目标总个数
+    ARIDLL_OBJINFO  stAlarmObjs[ST_OBJ_NUM];            //检测目标信息数组
+
+    //*****目标跟踪*****(长时航迹点，第0个为主目标送伺服跟踪)
+    int nTrackObjCnts;                                  //跟踪目标个数
+    ARIDLL_OBJINFO  stTrackers[LT_OBJ_NUM];             //跟踪器输出数组
+
+    // AI跟踪器协同控制指令输出,用于控制端侧NPU程序
+    AIT_Command stAI_TkCmd;
+
+    //搜索区域(对空场景为导引区域)
+    RECT32S   rsRecaptureRion;
+
+    //指示轨迹预测模块是否相信当前跟踪点
+    BBOOL bPredictJam;
+
+    //调试信息
+    ARIDLL_DEBUG_OUTPUT stDebugInfo;
+
+}ARIDLL_OUTPUT;
+
+/*************************************
+* 函数名称:            STD_CreatEOArithHandle()
+* 功能描述:            算法句柄创建
+* 创建日期:            2025/6/24
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API ArithHandle：算法句柄
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API ArithHandle STD_CreatEOArithHandle();
+
+
+/*************************************
+* 函数名称:            STD_CreatEOArithNamedHandle()
+* 功能描述:            指定算法句柄名称的句柄创建，每个句柄创建都有自己独立的配置文件路径，避免设置重复参数，比如句柄名称
+                       实际上也可以允许句柄名称重复，不建议这样做，因为不能起到区分作用，参数设置请参考文件
+* 创建日期:            2025/8/8
+* 输入参数:            const char * configPath:算法创建相关配置json名称
+* 输出参数:            
+* 返回值:              STD_TRACKER_API ArithHandle:算法句柄
+* 调用关系:            
+* 其它说明:            
+*************************************/
+STD_TRACKER_API ArithHandle STD_CreatEOArithNamedHandle(const char* configPath);
+
+
+/*************************************
+* 函数名称:            STD_DeleteEOArithHandle()
+* 功能描述:            释放算法句柄
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArith：算法句柄
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void   STD_DeleteEOArithHandle(ArithHandle hArith);
+
+
+/*************************************
+* 函数名称:            ARIDLL_EOArithInit()
+* 功能描述:            执行算法模块初始化，默认为对空凝视场景
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArith：算法句柄
+* 输入参数:            int nWidth：输入图像宽度
+* 输入参数:            int nHeight：输入图像高度
+* 输入参数:            GD_PIXEL_FORMAT_E nPixelType:图像的像素类型
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            GLB_PT_TYPE nPixelType 暂未使用的参数
+*************************************/
+STD_TRACKER_API void   ARIDLL_EOArithInit(ArithHandle hArith, int nWidth, int nHeight, GD_PIXEL_FORMAT_E nPixelType);
+
+
+/*************************************
+* 函数名称:            ARIDLL_EOArithInitWithMode()
+* 功能描述:            执行算法模块初始化2 - 带模式的初始化，可指定系统模式和场景模式
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArith：算法句柄
+* 输入参数:            int nWidth：输入图像宽度
+* 输入参数:            int nHeight：输入图像高度
+* 输入参数:            GD_PIXEL_FORMAT_E nPixelType:图像的像素类型
+* 输入参数:            GLB_SYS_MODE nSysMode:系统模式
+* 输入参数:            GLB_SCEN_MODE nScenMode：场景模式
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            GLB_PT_TYPE nPixelType 暂未使用的参数
+*************************************/
+STD_TRACKER_API void   ARIDLL_EOArithInitWithMode(ArithHandle hArith, int nWidth, int nHeight, GD_PIXEL_FORMAT_E nPixelType,
+    GLB_SYS_MODE nSysMode, GLB_SCEN_MODE nScenMode);
+
+
+/*************************************
+* 函数名称:            ARIDLL_CreateAITracker()
+* 功能描述:            创建AI跟踪器模块
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArith：算法句柄
+* 输入参数:            int nWidth：输入图像宽度
+* 输入参数:            int nHeight：输入图像高度
+* 输入参数:            const char * configPath:AI跟踪参数配置文件名称
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API bool
+* 调用关系:            无
+* 其它说明:            创建AI跟踪器后，NeoTracker仅在切换为AI跟踪场景模式时才会使用AI跟踪
+*************************************/
+STD_TRACKER_API bool   ARIDLL_CreateAITracker(ArithHandle hArith, int nWidth, int nHeight, const char* configPath);
+
+
+/*************************************
+* 函数名称:            ARIDLL_RunController()
+* 功能描述:            目标搜跟流程，算法主处理逻辑
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输入参数:            GD_VIDEO_FRAME_S img ：输入图像
+* 输入参数:            ARIDLL_INPUTPARA stInputPara ：输入参数
+* 输入参数:            ARIDLL_OUTPUT * pstOutput ：输出结果
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API int ：主处理逻辑状态码
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API int ARIDLL_RunController(ArithHandle hArithSrc, GD_VIDEO_FRAME_S img, ARIDLL_INPUTPARA stInputPara, ARIDLL_OUTPUT* pstOutput);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SearchFrameTargets()
+* 功能描述:            执行单帧小面目标检测算法
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输入参数:            GD_VIDEO_FRAME_S img ：输入图像
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API int ：单帧小面检测算法个数
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API int ARIDLL_SearchFrameTargets(ArithHandle hArithSrc, GD_VIDEO_FRAME_S img);
+
+
+/*************************************
+* 函数名称:            ARIDLL_MergeAITargets()
+* 功能描述:            接收外部AI识别目标，在主循环调用前调用以传入外部目标
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输入参数:            AI_Target * aiDetectArray：输入AI目标检测数组
+* 输入参数:            int aiNum：输入AI目标检测个数
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API int
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API int ARIDLL_MergeAITargets(ArithHandle hArithSrc, AI_Target* aiDetectArray,int aiNum);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SendReIDToTargets()
+* 功能描述:            接收外部目标reID特征，将特征传入到算法内部
+* 创建日期:            2025/6/25
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输入参数:            ReIDFeature * pReidFeatures ：ReID特征数组
+* 输入参数:            int ReIDNums ：ReID特征个数
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API int
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+//STD_TRACKER_API int ARIDLL_SendReIDToTargets(ArithHandle hArithSrc, ReIDFeature* pReidFeatures, int ReIDNums);
+
+
+/*************************************
+* 函数名称:            ARIDLL_LockCommand()
+* 功能描述:            视场内自适应锁定，智能锁定，有目标锁目标，无目标根据项目需求执行锁定方式,中心点XY一定要传，宽高可选择
+*                        + 点选吸附
+*                        + 框选距离管道目标近距离吸附，远距离强锁框选区域
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            int nLockX:锁定点目标图像坐标中心点X
+* 输入参数:            int nLockY:锁定点目标图像坐标中心点Y
+* 输入参数:            int nLockW:锁定波门宽度, 目标宽度
+* 输入参数:            int nLockH:锁定波门高度, 目标高度
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            下发指令在下一帧转入锁定，宽高送0则算法根据识别结果自动选择波门
+*************************************/
+STD_TRACKER_API void ARIDLL_LockCommand(ArithHandle hArithSrc, int nLockX, int nLockY, int nLockW, int nLockH);
+
+
+/*************************************
+* 函数名称:            ARIDLL_LockCommand_DefaultSize()
+* 功能描述:            框选吸附
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            int nLockX:锁定点目标图像坐标中心点X
+* 输入参数:            int nLockY:锁定点目标图像坐标中心点Y
+* 输入参数:            int nLockW:锁定波门宽度, 目标宽度
+* 输入参数:            int nLockH:锁定波门高度, 目标高度
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_LockCommand_DefaultSize(ArithHandle hArithSrc, int nLockX, int nLockY, int nLockW, int nLockH);
+
+
+/*************************************
+* 函数名称:            ARIDLL_LockTargetByID()
+* 功能描述:            视场内根据批号锁定，当前帧处理
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            int nBatchID:锁定管道批号
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_LockTargetByID(ArithHandle hArithSrc, int nBatchID);
+
+
+/*************************************
+* 函数名称:            ARIDLL_GuideLockMultiCommand()
+* 功能描述:            视场外引导锁定-支持批量锁定
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输入参数:            TargetGuide * guideList
+*                       SINT32              ID;                 //目标批号，传递给锁定后目标(必填，0为无效值，外部指定跟踪目标输出批号)
+                        UBYTE8              bIsCoLocate;        //协同定位标记，直接透传到目标(非必须，默认0)
+                        Pole                stTargetPole;       //目标极坐标(伺服系目标方位、俯仰、目标测距)
+                        PointBLH            stTargetPos;        //目标GPS坐标(大地系纬经高)
+                        FLOAT32             fGuideAzSpeed;      //实际锁定点方位角速度(非必须，默认0)
+                        FLOAT32             fGuidePtSpeed;      //实际锁定点俯仰角速度(非必须，默认0)
+                        SINT32              nGuideFocus;        //引导时的焦距值(必填)
+                        SINT32              nMaxFocus;          //最大焦距值(必填)
+                        SINT32              nLockX;             //锁定点当前图像坐标X(图像系X)
+                        SINT32              nLockY;             //锁定点当前图像坐标Y(图像系Y)
+                        BBOOL               bInFOV;             //在视场判断(图像系目标在视场判断)
+                        stTargetPole、stTargetPos、nLockX(nLockY,bInFOV),三选一，当前建议选择stTargetPole
+* 输入参数:            int num:引导目标个数
+* 输入参数:            int nGuideAge:引导生命周期，帧数衡量，如50HZ平台，1s填50，外部设定
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            对空引导在焦距没有拉满的情况下，不要出现是小目标的情况
+*************************************/
+STD_TRACKER_API void ARIDLL_GuideLockMultiCommand(ArithHandle hArithSrc, TargetGuide* guideList, int num, int nGuideAge);
+
+
+/*************************************
+* 函数名称:            ARIDLL_unLockCommand()
+* 功能描述:            解锁
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_unLockCommand(ArithHandle hArithSrc);
+
+
+/*************************************
+* 函数名称:            ARIDLL_AdjustTrackRect()
+* 功能描述:            微调对地主跟踪器跟踪框
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输入参数:            int dx:X 方向的偏移量，向左为负；向右为正
+* 输入参数:            int dy:Y 方向的偏移量，向上为负；向下为正
+* 输入参数:            int dw ：目标宽度扩大，缩减大小
+* 输入参数:            int dh ：目标高度扩大，缩减大小
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_AdjustTrackRect(ArithHandle hArithSrc,int dx,int dy,int dw,int dh);
+
+
+/*************************************
+* 函数名称:            ARIDLL_ReadSetParamFile()
+* 功能描述:            读取序列化参数：从文件
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc：算法句柄
+* 输入参数:            const char * configFilePath：保存了算法参数的文件名
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API bool
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API bool ARIDLL_ReadSetParamFile(ArithHandle hArithSrc, const char* configFilePath);
+
+
+/*************************************
+* 函数名称:            ARIDLL_ReadSetParamStream()
+* 功能描述:            读取序列化参数：从buffer
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            const char * configsstream:保存了算法参数的文件流
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API bool : 设置成功标记
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API bool ARIDLL_ReadSetParamStream(ArithHandle hArithSrc,const char* configsstream);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SetSysMode()
+* 功能描述:            设置外部系统工作模式
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            GLB_SYS_MODE nSysMode:系统模式
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_SetSysMode(ArithHandle hArithSrc, GLB_SYS_MODE nSysMode);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SetScenMode()
+* 功能描述:            设置工作场景,对空，对地，对海等
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            GLB_SCEN_MODE nScenMode:工作场景
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_SetScenMode(ArithHandle hArithSrc, GLB_SCEN_MODE nScenMode);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SetForceMemTrack()
+* 功能描述:            设置强制记忆跟踪
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            int nMemCnt:强制进记忆帧数,超出强制记忆帧数则退出记忆
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_SetForceMemTrack(ArithHandle hArithSrc, int nMemCnt);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SetSkyLineCaliPoints()
+* 功能描述:            接收外部传入的天地线标定点，最多支持360个点
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            ANGLE32F * SkyLinePoints:天地线标定点数组首地址
+* 输入参数:            int N:天地线标定点个数，最多360个
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_SetSkyLineCaliPoints(ArithHandle hArithSrc, ANGLE32F* SkyLinePoints, int N);
+
+
+/*************************************
+* 函数名称:            ARIDLL_ExportParamFile()
+* 功能描述:            输出算法配置文件
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            const char * configFilePath:输出算法配置的文件名
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_ExportParamFile(ArithHandle hArithSrc, const char* configFilePath);
+
+
+
+/*************************************
+* 函数名称:            ARIDLL_ExportOSDJson()
+* 功能描述:            输出算法调试json字符串流，算法OSD叠加功能
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            char * Buffer:接收OSD流的缓冲区
+* 输入参数:            int bufferSize:接收OSD流的缓冲区的大小
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API int：成功返回0
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API int ARIDLL_ExportOSDJson(ArithHandle hArithSrc, char* Buffer, int bufferSize);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SetSOTRect()
+* 功能描述:            按照固定大小调整波门
+* 创建日期:            2025/6/24
+* 输入参数:            ArithHandle hArithSrc
+* 输入参数:            int objX:新的锁定X中心
+* 输入参数:            int objY:新的锁定Y中心
+* 输入参数:            int objW:新的锁定目标宽度
+* 输入参数:            int objH:新的锁定目标高度
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_SetSOTRect(ArithHandle hArithSrc, int objX, int objY, int objW, int objH);
+
+
+/*************************************
+* 函数名称:            ARIDLL_SetPredictInfo()
+* 功能描述:            外部预测的主跟踪位置点
+* 创建日期:            2025/7/7
+* 输入参数:            ArithHandle hArithSrc:算法句柄
+* 输入参数:            int X1:当前帧预测x位置
+* 输入参数:            int Y1:当前帧预测y位置
+* 输入参数:            int X2:8s后预测的x位置
+* 输入参数:            int Y2:8s后预测的y位置
+* 输出参数:            无
+* 返回值:              STD_TRACKER_API void
+* 调用关系:            无
+* 其它说明:            无
+*************************************/
+STD_TRACKER_API void ARIDLL_SetPredictInfo(ArithHandle hArithSrc, int X1, int Y1, int X2, int Y2);
+
+
+#ifdef __cplusplus
+}
+#endif
+
+
+
+
+#endif

tests/cpp/Test_GeoCorrect.cpp

@@ -0,0 +1,47 @@
+#include "API_GeoCorrect.h"
+#include "StitchStruct.h"
+#include "opencv2/opencv.hpp"
+
+
+int main(int argc, char** argv)
+{
+    printf("Test_GeoCorrect\n");
+    API_GeoCorrect* geoCorrect = API_GeoCorrect::Create();
+
+    FrameInfo info;
+    info.camInfo.fPixelSize = 4;
+    info.camInfo.nFocus = 48;
+    
+
+    info.servoInfo.fServoAz = 1;
+    info.servoInfo.fServoPt = -78;
+
+    info.nWidth = 1920;
+    info.nHeight = 1080;
+
+    info.craft.stPos.B = 30.0;
+    info.craft.stPos.L = 114.0;
+    info.craft.stPos.H = 1000.0;
+
+    info.craft.stAtt.fRoll = 1.0;
+    info.craft.stAtt.fPitch = -2.0;
+    info.craft.stAtt.fYaw = 134.0;
+
+    info.nEvHeight = 1000;
+
+
+
+    geoCorrect->Init(info);
+
+    info.craft.stPos.B += 0.1;
+    info.craft.stPos.L += 0.1;
+    info.craft.stPos.H += 100;
+
+    
+
+
+
+
+    geoCorrect->Correct(cv::Mat(), info);
+    return 0;
+}

tests/python/1.py

@@ -0,0 +1,78 @@
+import os
+import sys
+import ctypes
+from pathlib import Path
+import numpy as np
+
+# 获取项目根目录（当前文件所在目录的父目录）
+project_root = Path(__file__).parent.parent.parent
+bin_dir = project_root / "Bin"
+
+# 添加Bin目录到Python模块搜索路径
+if str(bin_dir) not in sys.path:
+    sys.path.insert(0, str(bin_dir))
+
+# 预加载依赖库，确保动态链接器能找到它们
+lib_guide_stitch = bin_dir / "libGuideStitch.so"
+if lib_guide_stitch.exists():
+    try:
+        ctypes.CDLL(str(lib_guide_stitch), mode=ctypes.RTLD_GLOBAL)
+    except Exception as e:
+        print(f"警告: 预加载libGuideStitch.so失败: {e}")
+
+# 导入模块
+from UStitcher import API_UnderStitch, FrameInfo
+
+import cv2
+
+
+frame_info = FrameInfo()
+
+
+
+frame_info.nFrmID = 1
+frame_info.craft.stPos.B = 39
+frame_info.craft.stPos.L = 120
+frame_info.craft.stPos.H = 1000
+frame_info.camInfo.nFocus = 40
+frame_info.camInfo.fPixelSize = 12
+
+frame_info.servoInfo.fServoAz = 90
+frame_info.servoInfo.fServoPt = -45
+
+frame_info.nEvHeight = 1200
+
+frame_info.nWidth = 1280
+frame_info.nHeight = 1024
+
+
+
+stitcher = API_UnderStitch.Create()
+
+
+# 先初始化（设置原点）
+pan_info = stitcher.Init(frame_info)
+print(f"初始化成功，全景图尺寸: {pan_info.m_pan_width} x {pan_info.m_pan_height}")
+
+def warpPointWithH(H, pt):
+    wp = H @ np.array([pt[0],pt[1],1]).T
+    return wp / wp[2]
+
+for i in range(100):
+    frame_info.nFrmID = i
+    frame_info.craft.stPos.B = 39
+    frame_info.craft.stPos.L = 120
+    frame_info.craft.stPos.H = 1000
+    frame_info.camInfo.nFocus = 40
+    frame_info.camInfo.fPixelSize = 12
+
+    frame_info.servoInfo.fServoAz = 90
+    frame_info.servoInfo.fServoPt = -45
+
+    H = stitcher.getHomography(frame_info)
+    print(f"单应性矩阵 H:\n{H}")
+
+    wp = warpPointWithH(H, np.array([100,111]))
+    print(f"物方坐标:\n{wp}")
+
+print("done")

tests/python/ProcDJ.py

@@ -0,0 +1,427 @@
+import os
+import sys
+import ctypes
+import re
+from pathlib import Path
+
+import numpy as np
+from dataclasses import dataclass
+from typing import List, Tuple
+
+
+sys.path.append("/media/wang/WORK/wangchongwu_gitea_2023/StitchVideo/Bin")
+
+
+# 导入模块
+from UStitcher import API_UnderStitch, FrameInfo, UPanInfo, UPanConfig
+
+
+import cv2
+
+@dataclass
+class FrameData:
+    """帧数据"""
+    frame_number: int = 0
+    time_range: str = ""
+    frame_cnt: int = 0
+    timestamp: str = ""
+    focal_len: float = 0.0
+    dzoom_ratio: float = 0.0
+    latitude: float = 0.0
+    longitude: float = 0.0
+    rel_alt: float = 0.0
+    abs_alt: float = 0.0
+    gb_yaw: float = 0.0
+    gb_pitch: float = 0.0
+    gb_roll: float = 0.0
+    real_focal_mm: float = 0.0
+    pixel_size_um: float = 0.0
+
+
+def extract_value_from_brackets(line: str, key: str) -> float:
+    """从方括号格式中提取值，例如 [key: value] 或 [key: value other]"""
+    # 匹配模式1：单独的方括号 [key: value]
+    pattern1 = rf'\[{re.escape(key)}:\s*([^\]]+?)\]'
+    match = re.search(pattern1, line)
+    if match:
+        value_str = match.group(1).strip()
+        # 提取第一个数值（可能后面有其他数据）
+        num_match = re.search(r'-?\d+\.?\d*', value_str)
+        if num_match:
+            try:
+                return float(num_match.group())
+            except ValueError:
+                pass
+    
+    # 匹配模式2：在方括号内的键值对，例如 [rel_alt: 300.030 abs_alt: 314.064]
+    # 匹配 [xxx key: value xxx] 格式
+    pattern2 = rf'\[[^\]]*?{re.escape(key)}:\s*([^\s\]]+?)(?:\s|])'
+    match = re.search(pattern2, line)
+    if match:
+        value_str = match.group(1).strip()
+        try:
+            return float(value_str)
+        except ValueError:
+            pass
+    
+    return None  # 使用None表示未找到
+
+def extract_value_after(line: str, key: str) -> float:
+    """从字符串中提取指定键后面的值（兼容方括号和非方括号格式）"""
+    # 先尝试方括号格式
+    value = extract_value_from_brackets(line, key)
+    if value is not None:
+        return value
+    
+    # 如果没有方括号，尝试直接查找格式
+    pos = line.find(key)
+    if pos == -1:
+        return 0.0
+    pos += len(key)
+    # 跳过可能的空格和冒号
+    while pos < len(line) and line[pos] in ' :':
+        pos += 1
+    # 找到值的结束位置（空格、逗号、换行等）
+    end = pos
+    while end < len(line) and line[end] not in ' ,\n\r\t]':
+        end += 1
+    if end == pos:
+        return 0.0
+    try:
+        return float(line[pos:end])
+    except ValueError:
+        return 0.0
+
+
+def infer_camera_params_h30(frame: FrameData, filename: str) -> None:
+    """获取真实焦距和像元尺寸 - H30版本"""
+    if "_W" in filename:
+        frame.real_focal_mm = 6.72
+        frame.pixel_size_um = 2.4
+    elif "_Z" in filename:
+        frame.real_focal_mm = 6.72 * 2
+        frame.pixel_size_um = 2.4
+    elif "_T" in filename:
+        frame.real_focal_mm = 24.0
+        frame.pixel_size_um = 12.0
+    # else: 保持默认值
+
+
+def parse_dji_srt(filename: str) -> List[FrameData]:
+    """解析DJI SRT文件"""
+    frames = []
+    
+    try:
+        with open(filename, 'r', encoding='utf-8') as file:
+            lines = []
+            for line in file:
+                line_stripped = line.rstrip('\n\r')
+                lines.append(line_stripped)
+                
+                # 每5行一组（包括空行）
+                if len(lines) >= 5:
+                    frame = FrameData()
+                    # Line 1: Frame number
+                    try:
+                        frame.frame_number = int(lines[0])
+                    except ValueError:
+                        lines = []
+                        continue
+                    
+                    # Line 2: Time range
+                    frame.time_range = lines[1]
+                    
+                    # Line 3: FrameCnt and timestamp
+                    parts = lines[2].split()
+                    if len(parts) >= 3:
+                        try:
+                            frame.frame_cnt = int(parts[1])
+                            frame.timestamp = parts[2]
+                        except (ValueError, IndexError):
+                            pass
+                    
+                    # Line 4: Metadata
+                    meta = lines[3]
+                    # 使用改进的解析函数，支持方括号格式
+                    frame.focal_len = extract_value_after(meta, "focal_len")
+                    frame.dzoom_ratio = extract_value_after(meta, "dzoom_ratio")
+                    frame.latitude = extract_value_after(meta, "latitude")
+                    frame.longitude = extract_value_after(meta, "longitude")
+                    frame.rel_alt = extract_value_after(meta, "rel_alt")
+                    frame.abs_alt = extract_value_after(meta, "abs_alt")
+                    frame.gb_yaw = extract_value_after(meta, "gb_yaw")
+                    frame.gb_pitch = extract_value_after(meta, "gb_pitch")
+                    frame.gb_roll = extract_value_after(meta, "gb_roll")
+                    
+                    # 调试：打印前几帧的meta信息和解析结果
+                    if frame.frame_number <= 3:
+                        print(f"Frame {frame.frame_number} meta: {meta}")
+                        print(f"  解析结果: lat={frame.latitude}, lon={frame.longitude}, "
+                              f"alt={frame.abs_alt}, yaw={frame.gb_yaw}, pitch={frame.gb_pitch}, roll={frame.gb_roll}")
+                    
+                    infer_camera_params_h30(frame, filename)
+                    frames.append(frame)
+                    
+                    # 清空lines，准备下一组
+                    lines = []
+                
+    except Exception as e:
+        print(f"错误: 无法打开文件 {filename}: {e}")
+        return frames
+    
+    return frames
+
+
+def proc_dj_video(video_path_list: List[str], srt_path_list: List[str], 
+                  cache_dir: str = "./cache", output_dir: str = "./google_tiles"):
+    """处理DJI视频"""
+    # 创建拼接器
+    stitcher = API_UnderStitch.Create(cache_dir)
+    stitcher.SetOutput("DJI", output_dir)
+    
+    # 打开第一个视频获取属性
+    cap = cv2.VideoCapture(video_path_list[0])
+    if not cap.isOpened():
+        print(f"错误: 无法打开视频 {video_path_list[0]}")
+        return
+    
+    fps = cap.get(cv2.CAP_PROP_FPS)
+    width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
+    height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
+    frame_count = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
+    
+    # 降采样
+    n_down_sample = 1
+    if width > 3000:
+        n_down_sample = 2
+    
+    # 解析SRT文件
+    srt_init = parse_dji_srt(srt_path_list[0])
+    if not srt_init:
+        print("错误: SRT文件解析失败")
+        cap.release()
+        return
+    
+    n_start = min(2000, len(srt_init) - 1)
+    
+    # 初始化FrameInfo
+    frame_info = FrameInfo()
+    frame_info.nFrmID = n_start
+    frame_info.camInfo.nFocus = srt_init[n_start].real_focal_mm
+    frame_info.camInfo.fPixelSize = srt_init[n_start].pixel_size_um * n_down_sample
+    
+    frame_info.craft.stAtt.fYaw = srt_init[n_start].gb_yaw
+    frame_info.craft.stAtt.fPitch = 0.0
+    frame_info.craft.stAtt.fRoll = srt_init[n_start].gb_roll
+    
+    frame_info.craft.stPos.B = srt_init[n_start].latitude
+    frame_info.craft.stPos.L = srt_init[n_start].longitude
+    frame_info.craft.stPos.H = srt_init[n_start].abs_alt
+    
+    frame_info.nEvHeight = int(srt_init[n_start].rel_alt)
+    
+    frame_info.servoInfo.fServoAz = 0.0
+    frame_info.servoInfo.fServoPt = srt_init[n_start].gb_pitch
+    
+    frame_info.nWidth = width // n_down_sample
+    frame_info.nHeight = height // n_down_sample
+    
+    # 初始化拼接器
+    print("初始化拼接器...")
+    pan_info = stitcher.Init(frame_info)
+    print(f"初始化成功，全景图尺寸: {pan_info.m_pan_width} x {pan_info.m_pan_height}")
+    
+    # 设置配置
+    config = UPanConfig()
+    config.bOutFrameTile = False
+    config.bOutGoogleTile = True
+    config.bUseBA = True
+    stitcher.SetConfig(config)
+    
+    # 获取全景图
+    mat_pan = stitcher.ExportPanMat()
+    
+    cap.release()
+    
+    # 创建输出视频
+    output_width = mat_pan.shape[1] // 8
+    output_height = mat_pan.shape[0] // 8
+    output_path = "DJ_stitchVL.mp4"
+    
+    # 尝试多个编码器，按优先级顺序
+    codecs_to_try = [
+        ('H264', 'H264'),
+        ('mp4v', 'mp4v'),
+        ('XVID', 'XVID'),
+        ('MJPG', 'MJPG'),
+    ]
+    
+    output = None
+    used_codec = None
+    for codec_name, fourcc_str in codecs_to_try:
+        try:
+            fourcc = cv2.VideoWriter_fourcc(*fourcc_str)
+            output = cv2.VideoWriter(output_path, fourcc, 5.0, (output_width, output_height), True)
+            if output.isOpened():
+                used_codec = codec_name
+                print(f"成功创建输出视频，使用编码器: {codec_name}")
+                break
+            else:
+                output.release()
+                output = None
+        except Exception as e:
+            if output:
+                output.release()
+                output = None
+            print(f"尝试编码器 {codec_name} 失败: {e}")
+            continue
+    
+    if output is None or not output.isOpened():
+        print(f"错误: 无法创建输出视频 {output_path}")
+        print(f"尝试了所有编码器: {[c[0] for c in codecs_to_try]}")
+        print(f"输出尺寸: {output_width}x{output_height}")
+        print("提示: 可能需要安装视频编码库（如 ffmpeg）或使用其他编码器")
+        return
+    
+    # 处理每个视频
+    for vid_idx, video_path in enumerate(video_path_list):
+        print(f"处理 {video_path}")
+        cap = cv2.VideoCapture(video_path)
+        if not cap.isOpened():
+            print(f"错误: 无法打开视频 {video_path}")
+            continue
+        
+        srt = parse_dji_srt(srt_path_list[vid_idx])
+        if not srt:
+            print(f"错误: 无法解析SRT文件 {srt_path_list[vid_idx]}")
+            cap.release()
+            continue
+        
+        cap.set(cv2.CAP_PROP_POS_FRAMES, n_start)
+        frm_id = n_start
+        
+        while True:
+            ret, mat = cap.read()
+            if not ret or mat is None or mat.size == 0:
+                print("视频结束")
+                cap.release()
+                break
+            
+            frm_id += 1
+            
+            if frm_id < n_start:
+                continue
+            
+            if frm_id >= len(srt):
+                print(f"警告: 帧ID {frm_id} 超出SRT数据范围")
+                break
+            
+            # 降采样
+            mat_ds2 = cv2.resize(mat, (width // n_down_sample, height // n_down_sample))
+            
+            # 更新FrameInfo
+            frame_info = FrameInfo()
+            frame_info.nFrmID = frm_id
+            frame_info.camInfo.nFocus = srt[frm_id].real_focal_mm
+            frame_info.camInfo.fPixelSize = srt[frm_id].pixel_size_um * n_down_sample
+            
+            frame_info.craft.stAtt.fYaw = srt[frm_id].gb_yaw
+            frame_info.craft.stAtt.fPitch = 0.0
+            frame_info.craft.stAtt.fRoll = srt[frm_id].gb_roll
+            
+            frame_info.craft.stPos.B = srt[frm_id].latitude
+            frame_info.craft.stPos.L = srt[frm_id].longitude
+            frame_info.craft.stPos.H = srt[frm_id].abs_alt
+            
+            frame_info.nEvHeight = int(srt[frm_id].abs_alt)
+            
+            frame_info.servoInfo.fServoAz = 0.0
+            frame_info.servoInfo.fServoPt = srt[frm_id].gb_pitch
+            
+            frame_info.nWidth = mat_ds2.shape[1]
+            frame_info.nHeight = mat_ds2.shape[0]
+            
+            # 每10帧处理一次
+            if frm_id % 10 != 0:
+                continue
+            
+            progress = float(frm_id) / frame_count * 100
+            
+            print(f"{progress:.1f}% B={frame_info.craft.stPos.B:.6f} L={frame_info.craft.stPos.L:.6f} "
+                  f"H={frame_info.craft.stPos.H:.2f} Yaw={frame_info.craft.stAtt.fYaw:.2f} "
+                  f"Pitch={frame_info.craft.stAtt.fPitch:.2f} Roll={frame_info.craft.stAtt.fRoll:.2f} "
+                  f"ServoAz={frame_info.servoInfo.fServoAz:.2f} ServoPt={frame_info.servoInfo.fServoPt:.2f}")
+            
+            # 运行拼接
+            import time
+            start_time = time.time()
+            stitcher.Run(mat_ds2, frame_info)
+            cost_time = time.time() - start_time
+            print(f"处理时间: {cost_time:.3f}秒")
+            
+            # 获取全景图并写入输出视频
+            mat_pan = stitcher.ExportPanMat()
+            if mat_pan is not None and mat_pan.size > 0:
+                # 转换为BGR
+                if len(mat_pan.shape) == 3 and mat_pan.shape[2] == 4:
+                    pan_rgb = cv2.cvtColor(mat_pan, cv2.COLOR_BGRA2BGR)
+                else:
+                    pan_rgb = mat_pan
+                
+                # 降采样
+                pan_rgb_ds = cv2.resize(pan_rgb, (output_width, output_height))
+                
+                # 写入视频
+                output.write(pan_rgb_ds)
+                
+                # 显示（可选）
+                cv2.imshow("pan_rgb", pan_rgb_ds)
+                if cv2.waitKey(1) & 0xFF == 27:  # ESC键退出
+                    break
+        
+        # 优化并输出当前全景图
+        print("优化并输出当前全景图...")
+        import time
+        start_time = time.time()
+        stitcher.OptAndOutCurrPan()
+        stitcher.Stop()
+        opt_time = time.time() - start_time
+        print(f"优化时间: {opt_time:.3f}秒")
+        
+        # 最终输出
+        mat_pan = stitcher.ExportPanMat()
+        if mat_pan is not None and mat_pan.size > 0:
+            if len(mat_pan.shape) == 3 and mat_pan.shape[2] == 4:
+                pan_rgb = cv2.cvtColor(mat_pan, cv2.COLOR_BGRA2BGR)
+            else:
+                pan_rgb = mat_pan
+            pan_rgb_ds = cv2.resize(pan_rgb, (output_width, output_height))
+            output.write(pan_rgb_ds)
+            cv2.imshow("pan_rgb", pan_rgb_ds)
+            cv2.waitKey(0)
+    
+    output.release()
+    cv2.destroyAllWindows()
+    print("处理完成")
+
+
+def main():
+    """主函数"""
+    video_path_list = []
+    srt_path_list = []
+    
+    # 修改为你的数据路径
+    folder = "/media/wang/data/K2D_data/"
+    
+    video_path_list.append(folder + "DJI_20250418153043_0006_W.MP4")
+    srt_path_list.append(folder + "DJI_20250418153043_0006_W.srt")
+    
+    # video_path_list.append(folder + "DJI_20250418153043_0006_W.MP4")
+    # srt_path_list.append(folder + "DJI_20250418153043_0006_W.srt")
+    
+    proc_dj_video(video_path_list, srt_path_list)
+
+
+if __name__ == "__main__":
+    main()
+

tests/python/UStitcher.pyi

@@ -0,0 +1,158 @@
+from __future__ import annotations
+
+from typing import Any, TypeAlias
+
+from numpy.typing import NDArray
+
+ImageLike: TypeAlias = NDArray[Any]
+
+
+class PointBLH:
+    """地理坐标系（单位：度）."""
+
+    B: float
+    L: float
+    H: float
+
+    def __init__(self) -> None: ...
+
+
+class EulerRPY:
+    """RPY 姿态角（单位：度）."""
+
+    fRoll: float
+    fPitch: float
+    fYaw: float
+
+    def __init__(self) -> None: ...
+
+
+class AirCraftInfo:
+    """载体信息."""
+
+    nPlaneID: int
+    stPos: PointBLH
+    stAtt: EulerRPY
+
+    def __init__(self) -> None: ...
+
+
+class CamInfo:
+    """相机信息."""
+
+    nFocus: int
+    fPixelSize: float
+    unVideoType: int
+    dCamx: float
+    dCamy: float
+    fAglReso: float
+
+    def __init__(self) -> None: ...
+
+
+class ServoInfo:
+    """伺服状态."""
+
+    fServoAz: float
+    fServoPt: float
+    fServoAzSpeed: float
+    fServoPtSpeed: float
+
+    def __init__(self) -> None: ...
+
+
+class FrameInfo:
+    """帧内外方位元素."""
+
+    nFrmID: int
+    craft: AirCraftInfo
+    camInfo: CamInfo
+    servoInfo: ServoInfo
+    nEvHeight: int
+    nWidth: int
+    nHeight: int
+
+    def __init__(self) -> None: ...
+
+
+class UPanInfo:
+    """下视全景图配置."""
+
+    m_pan_width: int
+    m_pan_height: int
+    scale: float
+    map_shiftX: float
+    map_shiftY: float
+
+    def __init__(self) -> None: ...
+
+
+class UPanConfig:
+    """下视拼接参数控制."""
+
+    bUseBA: bool
+    bOutFrameTile: bool
+    bOutGoogleTile: bool
+
+    def __init__(self) -> None: ...
+
+
+class API_UnderStitch:
+    """视频帧下视地理拼接."""
+
+    def Init(self, info: FrameInfo) -> UPanInfo:
+        """初始化拼接，返回全景图配置."""
+        ...
+
+    def SetOutput(self, name: str, outdir: str) -> None:
+        """配置输出标识和目录."""
+        ...
+
+    def Run(self, img: ImageLike, para: FrameInfo) -> int:
+        """运行拼接流程（cv::Mat/numpy.ndarray）."""
+        ...
+
+    def Stop(self) -> None:
+        """中止拼接流程."""
+        ...
+
+    def SetConfig(self, config: UPanConfig) -> None:
+        """更新运行参数."""
+        ...
+
+    def OptAndOutCurrPan(self) -> int:
+        """立即优化并输出当前全景图."""
+        ...
+
+    def ExportPanMat(self) -> ImageLike:
+        """获取当前全景图像."""
+        ...
+
+    def getHomography(self, info: FrameInfo) -> ImageLike:
+        """根据帧信息返回单应性矩阵."""
+        ...
+
+    @staticmethod
+    def Create(cachedir: str = "./cache") -> API_UnderStitch:
+        """创建 API_UnderStitch 实例."""
+        ...
+
+    @staticmethod
+    def Destroy(obj: API_UnderStitch) -> None:
+        """销毁 API_UnderStitch 实例."""
+        ...
+
+
+__all__ = [
+    "API_UnderStitch",
+    "AirCraftInfo",
+    "CamInfo",
+    "EulerRPY",
+    "FrameInfo",
+    "ImageLike",
+    "PointBLH",
+    "ServoInfo",
+    "UPanConfig",
+    "UPanInfo",
+]
+

tests/python/sim_scan.py

@@ -0,0 +1,565 @@
+#!/usr/bin/env python3
+"""
+Downward scan simulator that visualises the projected camera footprint
+on top of a 2D map background. Camera, flight and scan parameters are
+interactive so engineers can inspect coverage overlap in real time.
+"""
+
+from __future__ import annotations
+
+import argparse
+import ctypes
+import math
+import sys
+import time
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Iterable, List, Optional, Tuple
+
+
+import numpy as np
+
+# -----------------------------------------------------------------------------
+# Environment bootstrap (reuse logic from tests/python/1.py)
+# -----------------------------------------------------------------------------
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+BIN_DIR = PROJECT_ROOT / "Bin"
+
+if str(BIN_DIR) not in sys.path:
+    sys.path.insert(0, str(BIN_DIR))
+
+LIB_GUIDE = BIN_DIR / "libGuideStitch.so"
+if LIB_GUIDE.exists():
+    try:
+        ctypes.CDLL(str(LIB_GUIDE), mode=ctypes.RTLD_GLOBAL)
+    except OSError as exc:
+        print(f"[WARN] Failed to preload {LIB_GUIDE.name}: {exc}")
+
+from UStitcher import API_UnderStitch, FrameInfo, UPanConfig
+import cv2
+
+# -----------------------------------------------------------------------------
+# Utility dataclasses
+# -----------------------------------------------------------------------------
+@dataclass
+class CameraProfile:
+    width: int = 1920
+    height: int = 1080
+    focus_mm: float = 40.0
+    pixel_size_um: float = 4
+
+
+@dataclass
+class FlightProfile:
+    start_lat: float = 39.0
+    start_lon: float = 120.0
+    altitude_m: float = 1000.0
+    speed_mps: float = 70.0
+    heading_deg: float = 90.0  # 0 -> east, 90 -> north
+
+
+@dataclass
+class ScannerProfile:
+    base_az_deg: float = 90.0
+    base_pitch_deg: float = -45.0
+    az_min_deg: float = 60.0
+    az_max_deg: float = 120.0
+    pitch_min_deg: float = -80.0
+    pitch_max_deg: float = -10.0
+    az_max_speed_degps: float = 25.0
+    az_max_acc_degps2: float = 60.0
+    pitch_max_speed_degps: float = 20.0
+    pitch_max_acc_degps2: float = 50.0
+
+
+@dataclass
+class MapView:
+    background: Optional[Path]
+    meters_per_pixel: float = 2.0
+    trail_length: int = 200
+
+
+@dataclass
+class SimulationOptions:
+    duration_s: float = 60.0
+    frame_rate: float = 25.0
+    overlay_alpha: float = 0.25
+    display_scale: float = 0.03  # 直接作用在全景尺度，控制画布大小
+
+
+@dataclass
+class AxisState:
+    angle: float
+    velocity: float
+    phase: float = 0.0
+
+
+@dataclass
+class AxisProfile:
+    low: float
+    high: float
+    mid: float
+    amplitude: float
+    omega: float
+
+
+# -----------------------------------------------------------------------------
+# Helper utilities
+# -----------------------------------------------------------------------------
+def clamp(value: float, low: float, high: float) -> float:
+    if low > high:
+        low, high = high, low
+    return max(low, min(high, value))
+
+
+def make_aircraft_polygon(center: np.ndarray, heading_deg: float, size: float = 18.0) -> np.ndarray:
+    """Construct a simple aircraft-shaped polygon oriented by heading."""
+    base = np.array(
+        [
+            [size, 0.0],              # nose
+            [-0.4 * size, 0.35 * size],
+            [-0.1 * size, 0.0],
+            [-0.4 * size, -0.35 * size],
+        ],
+        dtype=np.float32,
+    )
+    rad = math.radians(heading_deg)
+    rot = np.array(
+        [
+            [math.cos(rad), -math.sin(rad)],
+            [math.sin(rad), math.cos(rad)],
+        ],
+        dtype=np.float32,
+    )
+    pts = base @ rot.T
+    pts[:, 1] *= -1  # screen Y axis points downward
+    pts += center.astype(np.float32)
+    return pts
+
+
+# -----------------------------------------------------------------------------
+# Map drawing helper
+# -----------------------------------------------------------------------------
+def _blank_canvas(height: int = 900, width: int = 900) -> np.ndarray:
+    canvas = np.full((height, width, 3), 75, dtype=np.uint8)
+    step = 100
+    for x in range(0, width, step):
+        cv2.line(canvas, (x, 0), (x, height - 1), (45, 45, 45), 1)
+    for y in range(0, height, step):
+        cv2.line(canvas, (0, y), (width - 1, y), (45, 45, 45), 1)
+    return canvas
+
+
+class MapCanvas:
+    def __init__(self, view: MapView):
+        self.view = view
+        if view.background and view.background.exists():
+            image = cv2.imread(str(view.background), cv2.IMREAD_COLOR)
+            if image is None:
+                print(f"[WARN] Failed to load map '{view.background}', using blank canvas.")
+                image = _blank_canvas()
+        else:
+            image = _blank_canvas()
+        self.base = image
+        self.origin = np.array([image.shape[1] // 2, image.shape[0] // 2], dtype=np.float32)
+        self.history: List[np.ndarray] = []
+
+    def geo_to_px(self, pts_en: np.ndarray) -> np.ndarray:
+        """Convert local east/north meters to pixel coordinates."""
+        scale = 1.0 / self.view.meters_per_pixel
+        px = np.empty_like(pts_en)
+        px[:, 0] = self.origin[0] + pts_en[:, 0] * scale
+        px[:, 1] = self.origin[1] - pts_en[:, 1] * scale
+        return px
+
+    def push_trail(self, polygon_px: np.ndarray) -> None:
+        self.history.append(polygon_px.astype(np.int32))
+        if len(self.history) > self.view.trail_length:
+            self.history = self.history[-self.view.trail_length :]
+
+    def draw(self, live_poly: np.ndarray, aircraft_px: Tuple[int, int], info: Iterable[str]) -> np.ndarray:
+        frame = self.base.copy()
+        overlay = frame.copy()
+        for idx, poly in enumerate(self.history):
+            alpha = max(0.05, 1.0 - (len(self.history) - idx) / len(self.history))
+            cv2.fillPoly(overlay, [poly], (30, 90, 180), lineType=cv2.LINE_AA)
+            cv2.addWeighted(overlay, alpha * 0.2, frame, 1 - alpha * 0.2, 0, frame)
+        cv2.polylines(frame, [live_poly.astype(np.int32)], True, (0, 255, 0), 2, cv2.LINE_AA)
+        cv2.circle(frame, aircraft_px, 6, (0, 0, 255), -1, cv2.LINE_AA)
+
+        y = 20
+        for line in info:
+            cv2.putText(frame, line, (10, y), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255, 255, 255), 1, cv2.LINE_AA)
+            y += 20
+        return frame
+
+
+# -----------------------------------------------------------------------------
+# Simulation core
+# -----------------------------------------------------------------------------
+def warpPointWithH(H: np.ndarray, pt: np.ndarray) -> np.ndarray:
+    """使用H矩阵投影点坐标（参考1.py的实现）"""
+    wp = H @ np.array([pt[0], pt[1], 1.0])
+    return wp / wp[2]
+
+
+class UnderStitchSimulator:
+    def __init__(
+        self,
+        camera: CameraProfile,
+        flight: FlightProfile,
+        scanner: ScannerProfile,
+        map_view: MapView,
+        options: SimulationOptions,
+    ):
+        self.camera = camera
+        self.flight = flight
+        self.scanner = scanner
+        self.map = MapCanvas(map_view)
+        self.options = options
+
+        if options.display_scale <= 0:
+            raise ValueError("display_scale must be positive.")
+
+        self.time_step = 1.0 / options.frame_rate
+        self.total_frames = int(options.duration_s * options.frame_rate)
+        self.stitcher = API_UnderStitch.Create(str((PROJECT_ROOT / "cache").resolve()))
+        self.frame_info = FrameInfo()
+        self._init_frame_info()
+        
+        # 初始化拼接器并获取全景图信息
+        self.pan_info = self.stitcher.Init(self.frame_info)
+        print(f"[INFO] 初始化成功，全景图尺寸: {self.pan_info.m_pan_width} x {self.pan_info.m_pan_height}")
+        
+        # 根据 display_scale 缩放全景参数，减小绘制画布
+        scale = self.options.display_scale
+        self.pan_info.scale = self.pan_info.scale * 0.5
+        self.scaled_pan_width = max(1, int(self.pan_info.m_pan_width * scale))
+        self.scaled_pan_height = max(1, int(self.pan_info.m_pan_height * scale))
+        self.scaled_pan_scale = self.pan_info.scale * scale
+        self.scaled_shift_x = self.pan_info.map_shiftX * scale
+        self.scaled_shift_y = self.pan_info.map_shiftY * scale
+        
+        # 构建从地理坐标到全景图的变换矩阵（参考Arith_UnderStitch.cpp的getAffineFromGeo2Pan）
+        self.H_geo2pan = np.array([
+            [self.scaled_pan_scale, 0, self.scaled_shift_x],
+            [0, -self.scaled_pan_scale, self.scaled_pan_height + self.scaled_shift_y],
+            [0, 0, 1]
+        ], dtype=np.float64)
+
+        # 初始化伺服扫描状态（正弦扇扫：边界速度最小、中心最大）
+        self.az_profile = self._build_axis_profile(
+            self.scanner.az_min_deg,
+            self.scanner.az_max_deg,
+            self.scanner.az_max_speed_degps,
+            self.scanner.az_max_acc_degps2,
+        )
+        self.pitch_profile = self._build_axis_profile(
+            self.scanner.pitch_min_deg,
+            self.scanner.pitch_max_deg,
+            self.scanner.pitch_max_speed_degps,
+            self.scanner.pitch_max_acc_degps2,
+        )
+        self.az_state = self._init_axis_state(self.az_profile, self.scanner.base_az_deg)
+        self.pitch_state = self._init_axis_state(self.pitch_profile, self.scanner.base_pitch_deg)
+
+        self.frame_info.servoInfo.fServoAz = self.az_state.angle
+        self.frame_info.servoInfo.fServoPt = self.pitch_state.angle
+
+    def _build_axis_profile(
+        self,
+        min_deg: float,
+        max_deg: float,
+        max_speed: float,
+        max_acc: float,
+    ) -> AxisProfile:
+        low, high = (min_deg, max_deg) if min_deg <= max_deg else (max_deg, min_deg)
+        mid = (low + high) * 0.5
+        amplitude = (high - low) * 0.5
+        if amplitude <= 1e-6:
+            return AxisProfile(low, high, mid, 0.0, 0.0)
+
+        omega_candidates = []
+        if max_speed > 0:
+            omega_candidates.append(max_speed / amplitude)
+        if max_acc > 0:
+            omega_candidates.append(math.sqrt(max_acc / amplitude))
+
+        omega = min(omega_candidates) if omega_candidates else 0.0
+        return AxisProfile(low, high, mid, amplitude, omega)
+
+    def _init_axis_state(self, profile: AxisProfile, base_angle: float) -> AxisState:
+        angle = clamp(base_angle, profile.low, profile.high)
+        if profile.amplitude <= 1e-6 or profile.omega <= 0.0:
+            return AxisState(angle=angle, velocity=0.0, phase=0.0)
+
+        rel = clamp((angle - profile.mid) / profile.amplitude, -1.0, 1.0)
+        phase = math.asin(rel)
+        velocity = profile.amplitude * profile.omega * math.cos(phase)
+        return AxisState(angle=angle, velocity=velocity, phase=phase)
+
+    def _init_frame_info(self) -> None:
+        fi = self.frame_info
+        fi.camInfo.nFocus = int(self.camera.focus_mm)
+        fi.camInfo.fPixelSize = float(self.camera.pixel_size_um)
+        fi.nWidth = self.camera.width
+        fi.nHeight = self.camera.height
+        fi.nEvHeight = int(self.flight.altitude_m)
+        fi.craft.stPos.B = self.flight.start_lat
+        fi.craft.stPos.L = self.flight.start_lon
+        fi.craft.stPos.H = self.flight.altitude_m
+        fi.servoInfo.fServoAz = self.scanner.base_az_deg
+        fi.servoInfo.fServoPt = self.scanner.base_pitch_deg
+
+    def _update_pose(self, step_idx: int) -> Tuple[float, float]:
+        t = step_idx * self.time_step
+        distance = self.flight.speed_mps * t
+        heading_rad = math.radians(self.flight.heading_deg)
+        east = distance * math.cos(heading_rad)
+        north = distance * math.sin(heading_rad)
+
+        lat = self.flight.start_lat + north / 111320.0
+        lon = self.flight.start_lon + east / (111320.0 * math.cos(math.radians(self.flight.start_lat)))
+        self.frame_info.craft.stPos.B = lat
+        self.frame_info.craft.stPos.L = lon
+        self.frame_info.nFrmID = step_idx
+
+        self.frame_info.craft.stAtt.fYaw = self.flight.heading_deg
+        self.frame_info.craft.stAtt.fPitch = 0.0
+        self.frame_info.craft.stAtt.fRoll = 0.0
+        return east, north
+
+    def _step_axis(self, state: AxisState, profile: AxisProfile, dt: float) -> None:
+        if profile.amplitude <= 1e-6 or profile.omega <= 0.0:
+            state.angle = profile.mid
+            state.velocity = 0.0
+            state.phase = 0.0
+            return
+
+        state.phase += profile.omega * dt
+        # Wrap phase to [-pi, pi] for numerical stability
+        state.phase = (state.phase + math.pi) % (2 * math.pi) - math.pi
+        state.angle = profile.mid + profile.amplitude * math.sin(state.phase)
+        state.velocity = profile.amplitude * profile.omega * math.cos(state.phase)
+
+    def _update_scanner(self, dt: float) -> None:
+        self._step_axis(self.az_state, self.az_profile, dt)
+        self._step_axis(self.pitch_state, self.pitch_profile, dt)
+
+        self.frame_info.servoInfo.fServoAz = self.az_state.angle
+        self.frame_info.servoInfo.fServoPt = self.pitch_state.angle
+
+    def _image_corners_to_geo(self, H_img2geo: np.ndarray) -> Optional[np.ndarray]:
+        if H_img2geo is None or H_img2geo.size != 9:
+            return None
+
+        img_corners = np.array(
+            [
+                [0.0, 0.0],
+                [self.camera.width, 0.0],
+                [self.camera.width, self.camera.height],
+                [0.0, self.camera.height],
+            ],
+            dtype=np.float64,
+        )
+
+        geo_corners = [warpPointWithH(H_img2geo, pt)[:2] for pt in img_corners]
+        return np.array(geo_corners, dtype=np.float64)
+
+    def _image_point_to_geo(self, H_img2geo: np.ndarray, pt: np.ndarray) -> Optional[np.ndarray]:
+        if H_img2geo is None or H_img2geo.size != 9:
+            return None
+        return warpPointWithH(H_img2geo, pt)[:2]
+
+    def _geo_to_pan(self, geo_pts: np.ndarray) -> np.ndarray:
+        pan_pts = [warpPointWithH(self.H_geo2pan, pt)[:2] for pt in geo_pts]
+        return np.array(pan_pts, dtype=np.float64)
+
+    def _compute_aircraft_geo_downward(self) -> Optional[np.ndarray]:
+        """
+        将俯仰角临时设置为 -90°，用于计算机体正下方在大地坐标中的投影。
+        """
+        original_pitch = self.frame_info.servoInfo.fServoPt
+        try:
+            self.frame_info.servoInfo.fServoPt = -90.0
+            H_down = self.stitcher.getHomography(self.frame_info)
+        finally:
+            self.frame_info.servoInfo.fServoPt = original_pitch
+
+        if H_down is None or H_down.size != 9:
+            return None
+
+        center_px = np.array([self.camera.width * 0.5, self.camera.height * 0.5], dtype=np.float64)
+        return self._image_point_to_geo(H_down, center_px)
+
+    def run(self) -> None:
+        print(f"[INFO] Running simulator for {self.total_frames} frames.")
+        t0 = time.time()
+        
+        # 创建全景图画布
+        pan_canvas = np.zeros((self.scaled_pan_height, self.scaled_pan_width, 3), dtype=np.uint8)
+        
+        try:
+            for idx in range(self.total_frames):
+
+                start = time.perf_counter()
+                sim_time = idx * self.time_step
+                east, north = self._update_pose(idx)
+                self._update_scanner(self.time_step)
+
+                H_img2geo = self.stitcher.getHomography(self.frame_info)
+                geo_footprint = self._image_corners_to_geo(H_img2geo)
+                if geo_footprint is None:
+                    print("[WARN] Image->geo projection failed; skipping frame.")
+                    continue
+
+                plane_pan = None
+                plane_geo_down = self._compute_aircraft_geo_downward()
+                if plane_geo_down is not None:
+                    plane_pan = self._geo_to_pan(np.array([plane_geo_down]))[0]
+
+                # 将当前覆盖区域投影到全景图
+                footprint_pan = self._geo_to_pan(geo_footprint)
+
+                # 累积绘制到原始全景图上，实现自然叠加效果
+                cv2.fillPoly(pan_canvas, [footprint_pan.astype(np.int32)], (30, 90, 180), lineType=cv2.LINE_AA)
+
+                # 基于当前累积结果生成可渲染帧
+                frame = pan_canvas.copy()
+                
+                # 绘制当前覆盖框（绿色边框）
+                cv2.polylines(frame, [footprint_pan.astype(np.int32)], True, (0, 255, 0), 2, cv2.LINE_AA)
+
+                # 飞机位置画红色三角形
+                if plane_pan is not None:
+                    # 根据全景图比例计算三角形大小
+                    triangle_size = max(15, int(50 / max(self.scaled_pan_scale, 0.01)))
+                    heading_rad = math.radians(self.flight.heading_deg)
+                    
+                    # 定义指向上方的三角形（航向方向）
+                    # 顶点在航向方向，底边垂直于航向
+                    top = np.array([0, -triangle_size], dtype=np.float32)
+                    left = np.array([-triangle_size * 0.6, triangle_size * 0.4], dtype=np.float32)
+                    right = np.array([triangle_size * 0.6, triangle_size * 0.4], dtype=np.float32)
+                    
+                    # 旋转三角形以匹配航向
+                    cos_h = math.cos(heading_rad)
+                    sin_h = math.sin(heading_rad)
+                    rot_mat = np.array([[cos_h, -sin_h], [sin_h, cos_h]], dtype=np.float32)
+                    
+                    top_rot = top @ rot_mat.T
+                    left_rot = left @ rot_mat.T
+                    right_rot = right @ rot_mat.T
+                    
+                    # 转换到像素坐标
+                    center = plane_pan.astype(np.float32)
+                    triangle_pts = np.array([
+                        (center + top_rot).astype(np.int32),
+                        (center + left_rot).astype(np.int32),
+                        (center + right_rot).astype(np.int32)
+                    ])
+                    
+                    cv2.fillPoly(frame, [triangle_pts], (0, 0, 255), lineType=cv2.LINE_AA)
+                    cv2.polylines(frame, [triangle_pts], True, (255, 255, 255), 1, cv2.LINE_AA)
+                
+                
+                # 添加HUD信息
+                hud = [
+                    f"Frame: {idx}/{self.total_frames}",
+                    f"Pan Size (orig): {self.pan_info.m_pan_width} x {self.pan_info.m_pan_height}",
+                    f"Pan Size (scaled): {self.scaled_pan_width} x {self.scaled_pan_height}",
+                    f"Pan Scale (orig/scaled): {self.pan_info.scale:.4f}/{self.scaled_pan_scale:.4f}",
+                    f"Lat/Lon: {self.frame_info.craft.stPos.B:.6f}, {self.frame_info.craft.stPos.L:.6f}",
+                    f"Altitude: {self.flight.altitude_m:.1f} m",
+                    f"Speed: {self.flight.speed_mps:.1f} m/s",
+                    f"Az/Pt: {self.frame_info.servoInfo.fServoAz:.1f}/{self.frame_info.servoInfo.fServoPt:.1f}",
+                ]
+                
+                y = 20
+                for line in hud:
+                    cv2.putText(frame, line, (10, y), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255, 255, 255), 1, cv2.LINE_AA)
+                    y += 20
+
+                display_frame = frame
+                if not math.isclose(self.options.display_scale, 1.0):
+                    display_frame = cv2.resize(
+                        frame,
+                        (0, 0),
+                        fx=self.options.display_scale,
+                        fy=self.options.display_scale,
+                        interpolation=cv2.INTER_AREA,
+                    )
+
+                cv2.imshow("UnderStitch Scan Simulator - Panorama View", display_frame)
+                key = cv2.waitKey(1)
+                if key == 27 or key == ord("q"):
+                    break
+                end = time.perf_counter()
+                print(f"frame {idx} took {(end-start)*1000:.1f} ms")
+                # remaining = max(0.0, t0 + sim_time + self.time_step - time.time())
+                # time.sleep(remaining)
+        finally:
+            API_UnderStitch.Destroy(self.stitcher)
+            cv2.destroyAllWindows()
+
+
+# -----------------------------------------------------------------------------
+# CLI
+# -----------------------------------------------------------------------------
+def parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(description="UnderStitch downward scan footprint simulator.")
+    parser.add_argument("--duration", type=float, default=60.0, help="Simulation duration in seconds.")
+    parser.add_argument("--fps", type=float, default=50.0, help="Output/solver frame rate.")
+    parser.add_argument("--speed", type=float, default=50.0, help="Aircraft speed (m/s).")
+    parser.add_argument("--heading", type=float, default=36.0, help="Aircraft heading (deg, 0=E).")
+    parser.add_argument("--alt", type=float, default=1200.0, help="Altitude above ground (m).")
+    parser.add_argument("--map", type=Path, default=None, help="Optional background map image.")
+    parser.add_argument("--scale", type=float, default=2.0, help="Meters per pixel for map rendering.")
+    parser.add_argument("--trail", type=int, default=200, help="Stored footprint polygons.")
+    parser.add_argument("--base_az", type=float, default=90.0, help="Center azimuth (deg).")
+    parser.add_argument("--base_pitch", type=float, default=-90.0, help="Base pitch (deg).")
+    parser.add_argument("--az_min", type=float, default=90, help="Minimum azimuth limit (deg).")
+    parser.add_argument("--az_max", type=float, default=90, help="Maximum azimuth limit (deg).")
+    parser.add_argument("--az_speed", type=float, default=0, help="Azimuth max speed (deg/s).")
+    parser.add_argument("--az_acc", type=float, default=0, help="Azimuth max acceleration (deg/s^2).")
+    parser.add_argument("--pitch_min", type=float, default=-45.0, help="Minimum pitch limit (deg).")
+    parser.add_argument("--pitch_max", type=float, default=45.0, help="Maximum pitch limit (deg).")
+    parser.add_argument("--pitch_speed", type=float, default=45.0, help="Pitch max speed (deg/s).")
+    parser.add_argument("--pitch_acc", type=float, default=20.0, help="Pitch max acceleration (deg/s^2).")
+    parser.add_argument("--display_scale", type=float, default=0.3, help="Panorama display scale factor.")
+    return parser.parse_args()
+
+
+def main() -> None:
+    args = parse_args()
+    camera = CameraProfile()
+    flight = FlightProfile(
+        altitude_m=args.alt,
+        speed_mps=args.speed,
+        heading_deg=args.heading,
+    )
+    scanner = ScannerProfile(
+        base_az_deg=args.base_az,
+        base_pitch_deg=args.base_pitch,
+        az_min_deg=args.az_min,
+        az_max_deg=args.az_max,
+        pitch_min_deg=args.pitch_min,
+        pitch_max_deg=args.pitch_max,
+        az_max_speed_degps=args.az_speed,
+        az_max_acc_degps2=args.az_acc,
+        pitch_max_speed_degps=args.pitch_speed,
+        pitch_max_acc_degps2=args.pitch_acc,
+    )
+    map_view = MapView(background=args.map, meters_per_pixel=args.scale, trail_length=args.trail)
+    options = SimulationOptions(
+        duration_s=args.duration,
+        frame_rate=args.fps,
+        display_scale=args.display_scale,
+    )
+    simulator = UnderStitchSimulator(camera, flight, scanner, map_view, options)
+    simulator.run()
+
+
+if __name__ == "__main__":
+    main()
+