OpenCV4 + XMake

.gitignore

@@ -1,6 +1,9 @@
 cmake-build-debug
 build
 .idea
+.xmake
+.vscode/.cache
+.vscode/compile_commands.json
 Mark
 armor_box_photo
 tools/TrainCNN/.idea
@@ -9,4 +12,4 @@ tools/TrainCNN/__pycache__
 others/include/config/config.h
 others/MV-UB31-Group0.config
 .DS_Store
-video
+video

.vscode/settings.json

@@ -0,0 +1,8 @@
+{
+        "clangd.arguments": [
+        "--compile-commands-dir=${workspaceFolder}/.vscode",
+        "--completion-style=detailed",
+        "--query-driver=/usr/bin/g++,/usr/bin/gcc,/usr/bin/c++",
+        "--header-insertion=never"
+    ],
+}

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2019 xinyang
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -13,6 +13,12 @@
 
 **如有BUG或者想交流的朋友欢迎积极联系我们**
 
+**分享部分比赛时摄像头录制的视频:**
+
+链接: https://pan.baidu.com/s/1LwxEpeYYblX3cSzb59MTVg 提取码: 84ju 复制这段内容后打开百度网盘手机App，操作更方便哦
+
+---
+
 运行效果：自瞄帧率120（摄像头最大帧率）,识别距离根据环境不同大约8米左右(5mm焦距镜头)。
 
 ![front](https://github.com/lloi7/SJTU-RM-CV-2019/blob/master/picture0.png)
@@ -24,6 +30,8 @@
 | ---------------------------------------------------- | -------------------------------------------- | ------------------------------------------------------------ | ---------------------------------------------------------- |
 | IntelNUC<br />MindVision工业相机×１<br />USB转TTL×１ | Ubuntu18.04<br />Ubuntu16.04<br />Ｗindows10 | OpenCV3.4.5<br />OpenCV_contrib3.4.5<br />Eigen3<br />MindVision相机驱动 | Ubuntu18/16 : cmake3+gcc7+g++7 <br />Win10 : cmake3+VS2019 |
 
+MindVision相机型号：MV-UBS31GC
+
 **关于Windows环境下的运行支持，仅保证程序可以编译运行。对与部分辅助功能，如生成自启动脚本则不支持。**
 
 **实际装载在步兵和哨兵上的运行环境为Ubuntu18.04。**
@@ -60,6 +68,8 @@ sudo ./run
 
 ```./run --help```可以查看所有命令行参数及其作用。所有调试选项都集成到了命令行参数中。
 
+**不使用任何参数直接运行将没有任何图像显示。**
+
 需要调参的部分：主要需要根据车辆情况而调参的参数存放在others/include/config/setconfig.h中
 
 ### 3.工作条件
@@ -118,7 +128,7 @@ sudo ./run
 ### 1.自瞄装甲板识别方式
 
 ​    首先对图像进行通道拆分以及二值化操作，再进行开闭运算，通过边缘提取和条件限制得出可能为灯条的部分。再对所有可能的灯条进行两两匹配，根据形状大小特性进行筛选，得出可能为装甲板的候选区。然后把所有候选区交给分类器判断，得出真实的装甲板及其数字id。最后根据优先级选取最终击打目标以及后续处理。
-
+​     ![autoaim](https://github.com/lloi7/SJTU-RM-CV-2019/blob/master/自瞄流程图.png)
 ### 2.能量机关识别方式
 
 ​    首先对图像进行二值化操作，然后进行一定腐蚀和膨胀，通过边缘提取和条件限制得出待击打叶片（锤子形）。在待击打叶片范围内进一步用类似方法寻找目标装甲板和流动条，在二者连线上寻找中心的“R”。根据目标装甲板坐标和中心坐标计算极坐标系下的目标角度，进而预测待击打点的坐标（小符为装甲板本身，大符需要旋转）。最后将待击打点坐标和图像中心的差值转换为yaw和pitch轴角度，增加一环PID后发送给云台主控板。

armor/include/armor_finder/armor_finder.h

@@ -9,7 +9,7 @@
 #include <systime.h>
 #include <constants.h>
 #include <opencv2/core.hpp>
-#include <opencv2/tracking.hpp>
+#include <opencv2/tracking/tracking.hpp>
 #include <serial.h>
 #include <armor_finder/classifier/classifier.h>
 #include <additions.h>

armor/src/armor_finder/armor_finder.cpp

@@ -67,7 +67,7 @@ void ArmorFinder::run(cv::Mat &src) {
                 if ((target_box.rect & cv::Rect2d(0, 0, 640, 480)) == target_box.rect) { // 判断装甲板区域是否脱离图像区域
                     if (!classifier) {                                          /* 如果分类器不可用 */
                         cv::Mat roi = src(target_box.rect).clone(), roi_gray;  /* 就使用装甲区域亮点数判断是否跟丢 */
-                        cv::cvtColor(roi, roi_gray, CV_RGB2GRAY);
+                        cv::cvtColor(roi, roi_gray, cv::COLOR_RGB2GRAY);
                         cv::threshold(roi_gray, roi_gray, 180, 255, cv::THRESH_BINARY);
                         contour_area = cv::countNonZero(roi_gray);
                     }

armor/src/armor_finder/find/find_armor_box.cpp

@@ -156,6 +156,7 @@ bool ArmorFinder::findArmorBox(const cv::Mat &src, ArmorBox &box) {
         for (auto &one_box : armor_boxes) {
             if (one_box.id != 0) {
                 box = one_box;
+                break;
             }
         }
         if (save_labelled_boxes) {

armor/src/armor_finder/find/find_light_blobs.cpp

@@ -83,11 +83,11 @@ bool ArmorFinder::findLightBlobs(const cv::Mat &src, LightBlobs &light_blobs) {
     }else{
         light_threshold = 200;
     }
-    cv::threshold(color_channel, src_bin_light, light_threshold, 255, CV_THRESH_BINARY); // 二值化对应通道
+    cv::threshold(color_channel, src_bin_light, light_threshold, 255, cv::THRESH_BINARY); // 二值化对应通道
     if (src_bin_light.empty()) return false;
     imagePreProcess(src_bin_light);                                  // 开闭运算
 
-    cv::threshold(color_channel, src_bin_dim, 140, 255, CV_THRESH_BINARY); // 二值化对应通道
+    cv::threshold(color_channel, src_bin_dim, 140, 255, cv::THRESH_BINARY); // 二值化对应通道
     if (src_bin_dim.empty()) return false;
     imagePreProcess(src_bin_dim);                                  // 开闭运算
 
@@ -100,8 +100,8 @@ bool ArmorFinder::findLightBlobs(const cv::Mat &src, LightBlobs &light_blobs) {
     std::vector<std::vector<cv::Point>> light_contours_light, light_contours_dim;
     LightBlobs light_blobs_light, light_blobs_dim;
     std::vector<cv::Vec4i> hierarchy_light, hierarchy_dim;
-    cv::findContours(src_bin_light, light_contours_light, hierarchy_light, CV_RETR_CCOMP, CV_CHAIN_APPROX_NONE);
-    cv::findContours(src_bin_dim, light_contours_dim, hierarchy_dim, CV_RETR_CCOMP, CV_CHAIN_APPROX_NONE);
+    cv::findContours(src_bin_light, light_contours_light, hierarchy_light, cv::RETR_CCOMP, cv::CHAIN_APPROX_NONE);
+    cv::findContours(src_bin_dim, light_contours_dim, hierarchy_dim, cv::RETR_CCOMP, cv::CHAIN_APPROX_NONE);
     for (int i = 0; i < light_contours_light.size(); i++) {
         if (hierarchy_light[i][2] == -1) {
             cv::RotatedRect rect = cv::minAreaRect(light_contours_light[i]);

armor/src/armor_finder/tracking_state/tracking_state.cpp

@@ -8,13 +8,13 @@
 #include <show_images/show_images.h>
 
 bool ArmorFinder::stateTrackingTarget(cv::Mat &src) {
-    auto pos = target_box.rect;
+    cv::Rect pos(target_box.rect);
     if(!tracker->update(src, pos)){ // 使用KCFTracker进行追踪
         target_box = ArmorBox();
         LOGW("Track fail!");
         return false;
     }
-    if((pos & cv::Rect2d(0, 0, 640, 480)) != pos){
+    if((cv::Rect2d(pos) & cv::Rect2d(0, 0, 640, 480)) != cv::Rect2d(pos)){
         target_box = ArmorBox();
         LOGW("Track out range!");
         return false;
@@ -52,7 +52,7 @@ bool ArmorFinder::stateTrackingTarget(cv::Mat &src) {
             }
         }else{ //　分类器不可用，使用常规方法判断
             cv::Mat roi_gray;
-            cv::cvtColor(roi, roi_gray, CV_RGB2GRAY);
+            cv::cvtColor(roi, roi_gray, cv::COLOR_RGB2GRAY);
             cv::threshold(roi_gray, roi_gray, 180, 255, cv::THRESH_BINARY);
             contour_area = cv::countNonZero(roi_gray);
             if(abs(cv::countNonZero(roi_gray) - contour_area) > contour_area * 0.3){

energy/src/energy/find/energy_finder.cpp

@@ -21,12 +21,12 @@ int Energy::findFans(const cv::Mat &src) {
     static Mat src_bin;
     src_bin = src.clone();
     if (src.type() == CV_8UC3) {
-        cvtColor(src_bin, src_bin, CV_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
+        cvtColor(src_bin, src_bin, cv::COLOR_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
     }
     std::vector<vector<Point> > fan_contours;
     FanStruct(src_bin);//图像膨胀，防止图像断开并更方便寻找
     if (show_process)imshow("fan struct", src_bin);
-    findContours(src_bin, fan_contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
+    findContours(src_bin, fan_contours, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE);
 
     for (auto &fan_contour : fan_contours) {
         if (!isValidFanContour(src_bin, fan_contour)) {
@@ -54,16 +54,16 @@ int Energy::findArmors(const cv::Mat &src) {
     static Mat src_bin;
     src_bin = src.clone();
     if (src.type() == CV_8UC3) {
-        cvtColor(src_bin, src_bin, CV_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
+        cvtColor(src_bin, src_bin, cv::COLOR_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
     }
     std::vector<vector<Point> > armor_contours;
     std::vector<vector<Point> > armor_contours_external;//用总轮廓减去外轮廓，只保留内轮廓，除去流动条的影响。
 
     ArmorStruct(src_bin);//图像膨胀，防止图像断开并更方便寻找
-    findContours(src_bin, armor_contours, CV_RETR_LIST, CV_CHAIN_APPROX_NONE);
+    findContours(src_bin, armor_contours, cv::RETR_LIST, cv::CHAIN_APPROX_NONE);
     if (show_process)imshow("armor struct", src_bin);
 
-    findContours(src_bin, armor_contours_external, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
+    findContours(src_bin, armor_contours_external, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE);
     for (int i = 0; i < armor_contours_external.size(); i++)//去除外轮廓
     {
         unsigned long external_contour_size = armor_contours_external[i].size();
@@ -103,12 +103,12 @@ bool Energy::findCenterR(const cv::Mat &src) {
     static Mat src_bin;
     src_bin = src.clone();
     if (src.type() == CV_8UC3) {
-        cvtColor(src_bin, src_bin, CV_BGR2GRAY);
+        cvtColor(src_bin, src_bin, cv::COLOR_BGR2GRAY);
     }
     std::vector<vector<Point> > center_R_contours;
     CenterRStruct(src_bin);
     if (show_process)imshow("R struct", src_bin);
-    findContours(src_bin, center_R_contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
+    findContours(src_bin, center_R_contours, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE);
     for (auto &center_R_contour : center_R_contours) {
         if (!isValidCenterRContour(center_R_contour)) {
             continue;
@@ -139,13 +139,13 @@ bool Energy::findFlowStripFan(const cv::Mat &src) {
     src_bin = src.clone();
     src_copy = src.clone();
     if (src.type() == CV_8UC3) {
-        cvtColor(src_bin, src_bin, CV_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
+        cvtColor(src_bin, src_bin, cv::COLOR_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
     }
     std::vector<vector<Point> > flow_strip_fan_contours;
     FlowStripFanStruct(src_bin);//图像膨胀，防止图像断开并更方便寻找
     if (show_process)imshow("flow strip fan struct", src_bin);
 
-    findContours(src_bin, flow_strip_fan_contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
+    findContours(src_bin, flow_strip_fan_contours, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE);
     std::vector<cv::RotatedRect> candidate_flow_strip_fans;
 
     for (auto &flow_strip_fan_contour : flow_strip_fan_contours) {
@@ -175,7 +175,7 @@ bool Energy::findFlowStrip(const cv::Mat &src) {
 
     if (src_bin.type() == CV_8UC1) // 黑白图像
     {
-        cvtColor(src_bin, src_bin, COLOR_GRAY2RGB);
+        cvtColor(src_bin, src_bin, cv::COLOR_GRAY2RGB);
 
     }
     std::vector<cv::RotatedRect> candidate_target_armors = target_armors;
@@ -189,13 +189,13 @@ bool Energy::findFlowStrip(const cv::Mat &src) {
         }
     }
 
-    cvtColor(src_bin, src_bin, CV_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
+    cvtColor(src_bin, src_bin, cv::COLOR_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
 
     FlowStripStruct(src_bin);//图像膨胀，防止图像断开并更方便寻找
     if (show_process)imshow("flow strip struct", src_bin);
 
     std::vector<vector<Point> > flow_strip_contours;
-    findContours(src_bin, flow_strip_contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
+    findContours(src_bin, flow_strip_contours, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE);
 
     for (auto candidate_flow_strip_fan: flow_strip_fans) {
         for (auto &flow_strip_contour : flow_strip_contours) {
@@ -261,7 +261,7 @@ bool Energy::findFlowStripWeak(const cv::Mat &src) {
 
     if (src_bin.type() == CV_8UC1) // 黑白图像
     {
-        cvtColor(src_bin, src_bin, COLOR_GRAY2RGB);
+        cvtColor(src_bin, src_bin, cv::COLOR_GRAY2RGB);
 
     }
     std::vector<cv::RotatedRect> candidate_armors = armors;
@@ -275,13 +275,13 @@ bool Energy::findFlowStripWeak(const cv::Mat &src) {
         }
     }
 
-    cvtColor(src_bin, src_bin, CV_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
+    cvtColor(src_bin, src_bin, cv::COLOR_BGR2GRAY);//若读取三通道视频文件，需转换为单通道
 
     FlowStripStruct(src_bin);//图像膨胀，防止图像断开并更方便寻找
     if (show_process)imshow("weak struct", src_bin);
 
     std::vector<vector<Point> > flow_strip_contours;
-    findContours(src_bin, flow_strip_contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
+    findContours(src_bin, flow_strip_contours, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE);
 
     for (auto &flow_strip_contour : flow_strip_contours) {
         if (!isValidFlowStripContour(flow_strip_contour)) {
@@ -335,6 +335,4 @@ bool Energy::findCenterROI(const cv::Mat &src) {
                                  Size2f(length * 1.7, length * 1.7), -90);
     return true;
 
-}
-
-
+}

others/include/camera/camera_wrapper.h

@@ -6,7 +6,7 @@
 #define _CAMERA_WRAPPER_H_
 
 #include <additions.h>
-#include <opencv2/core/core.hpp>
+#include <opencv2/core.hpp>
 #include <camera/wrapper_head.h>
 
 #ifdef Windows
@@ -33,7 +33,7 @@ private:
     tSdkCameraCapbility tCapability;
     tSdkFrameHead frame_info;
     BYTE *pby_buffer;
-    IplImage* iplImage;
+    cv::Mat image_header;
     int channel;
 
     RoundQueue<cv::Mat, 2> src_queue;
@@ -52,4 +52,4 @@ public:
 };
 
 
-#endif /* _CAMERA_WRAPPER_H_ */
+#endif /* _CAMERA_WRAPPER_H_ */

others/include/log.h

@@ -32,6 +32,7 @@
 
 #include <stdio.h>
 #include <systime.h>
+#include <iostream>
 
 /************** Define the control code *************/
 #define START_CTR           "\033[0"

others/src/camera/camera_wrapper.cpp

@@ -17,7 +17,6 @@ CameraWrapper::CameraWrapper(int exposure, int gain, int camera_mode, const std:
         mode(camera_mode),
         camera_cnts(2),
         camera_status(-1),
-        iplImage(nullptr),
         rgb_buffer(nullptr),
         channel(3),
         gain(gain),
@@ -27,9 +26,11 @@ CameraWrapper::CameraWrapper(int exposure, int gain, int camera_mode, const std:
 void cameraCallback(CameraHandle hCamera, BYTE *pFrameBuffer, tSdkFrameHead* pFrameHead,PVOID pContext){
     CameraWrapper *c = (CameraWrapper*)pContext;
     CameraImageProcess(hCamera, pFrameBuffer, c->rgb_buffer, pFrameHead);
-    auto iplImage = cvCreateImageHeader(cvSize(pFrameHead->iWidth, pFrameHead->iHeight), IPL_DEPTH_8U, c->channel);
-    cvSetData(iplImage, c->rgb_buffer, pFrameHead->iWidth * c->channel);  //此处只是设置指针，无图像块数据拷贝，不需担心转换效率
-    c->src_queue.push(cv::cvarrToMat(iplImage).clone());
+    // 使用 cv::Mat 替代 IplImage
+    cv::Mat img(pFrameHead->iHeight, pFrameHead->iWidth, 
+                c->channel == 3 ? CV_8UC3 : CV_8UC1, 
+                c->rgb_buffer, pFrameHead->iWidth * c->channel);
+    c->src_queue.push(img.clone());
 }
 
 bool CameraWrapper::init() {
@@ -134,15 +135,9 @@ bool CameraWrapper::read(cv::Mat &src) {
 
 bool CameraWrapper::readRaw(cv::Mat &src) {
     if (CameraGetImageBuffer(h_camera, &frame_info, &pby_buffer, 500) == CAMERA_STATUS_SUCCESS) {
-        if (iplImage) {
-            cvReleaseImageHeader(&iplImage);
-        }
-
-        iplImage = cvCreateImageHeader(cvSize(frame_info.iWidth, frame_info.iHeight), IPL_DEPTH_8U, 1);
-
-        cvSetData(iplImage, pby_buffer, frame_info.iWidth);  //此处只是设置指针，无图像块数据拷贝，不需担心转换效率
-
-        src = cv::cvarrToMat(iplImage).clone();
+        // 使用 cv::Mat 替代 IplImage
+        cv::Mat img(frame_info.iHeight, frame_info.iWidth, CV_8UC1, pby_buffer, frame_info.iWidth);
+        src = img.clone();
 
         //在成功调用CameraGetImageBuffer后，必须调用CameraReleaseImageBuffer来释放获得的buffer。
         //否则再次调用CameraGetImageBuffer时，程序将被挂起一直阻塞，直到其他线程中调用CameraReleaseImageBuffer来释放了buffer
@@ -156,16 +151,13 @@ bool CameraWrapper::readRaw(cv::Mat &src) {
 }
 
 bool CameraWrapper::readProcessed(cv::Mat &src) {
-//	cerr << "Get-1" << endl;
     if (CameraGetImageBuffer(h_camera, &frame_info, &pby_buffer, 500) == CAMERA_STATUS_SUCCESS) {
         CameraImageProcess(h_camera, pby_buffer, rgb_buffer,
                            &frame_info);  // this function is super slow, better not to use it.
-        if (iplImage) {
-            cvReleaseImageHeader(&iplImage);
-        }
-        iplImage = cvCreateImageHeader(cvSize(frame_info.iWidth, frame_info.iHeight), IPL_DEPTH_8U, channel);
-        cvSetData(iplImage, rgb_buffer, frame_info.iWidth * channel);  //此处只是设置指针，无图像块数据拷贝，不需担心转换效率
-        src = cv::cvarrToMat(iplImage).clone();
+        // 使用 cv::Mat 替代 IplImage
+        int mat_type = (channel == 3) ? CV_8UC3 : CV_8UC1;
+        cv::Mat img(frame_info.iHeight, frame_info.iWidth, mat_type, rgb_buffer, frame_info.iWidth * channel);
+        src = img.clone();
         //在成功调用CameraGetImageBuffer后，必须调用CameraReleaseImageBuffer来释放获得的buffer。
         //否则再次调用CameraGetImageBuffer时，程序将被挂起一直阻塞，直到其他线程中调用CameraReleaseImageBuffer来释放了buffer
         CameraReleaseImageBuffer(h_camera, pby_buffer);
@@ -193,4 +185,4 @@ CameraWrapper::~CameraWrapper() {
     //注意，先反初始化后再free
     if (rgb_buffer != nullptr)
         free(rgb_buffer);
-}
+}

others/src/options.cpp

@@ -6,12 +6,13 @@
 #include <log.h>
 #include <cstring>
 #include <map>
+#include <string>
 
-bool show_armor_box = false;
+bool show_armor_box = true;
 bool show_armor_boxes = false;
 bool show_light_blobs = false;
 bool show_origin = false;
-bool run_with_camera = false;
+bool run_with_camera = true;
 bool save_video = false;
 bool wait_uart = false;
 bool save_labelled_boxes = false;

tools/tty-permission.sh

@@ -0,0 +1,3 @@
+#!/bin/sh
+
+sudo touch /etc/udev/rules.d/70-ttyusb.rules

xmake.lua

@@ -0,0 +1,122 @@
+-- 设置项目信息
+set_project("SJTU-RM-CV")
+set_version("1.0.0")
+set_languages("c++17")
+
+-- 设置构建模式
+set_rules("mode.release")
+set_optimize("fastest")
+
+--导出构建指令
+add_rules("plugin.compile_commands.autoupdate", {outputdir = ".vscode"})
+
+-- 定义宏
+add_defines('PATH=\"$(projectdir)\"')
+add_defines("Linux")
+
+-- 检查是否存在 config.h
+if os.isfile("others/include/config/config.h") or os.isfile("others/include/config.h") then
+    add_defines("WITH_CONFIG")
+    print("Found config.h")
+end
+
+-- 添加依赖
+add_requires("eigen", {system = false})
+add_requires("pthread")
+
+--- OpenCV 4.6.0 配置（启用 ffmpeg 和 contrib）
+add_requires("opencv", {
+    system = false,
+    configs = {
+        ffmpeg = false,
+        eigen = true,
+        png = true,
+        jpeg = true,
+        webp = false,
+        tiff = false,
+        quirc = false,
+        opengl = false,
+        protobuf = false,
+        bundled = true,
+        gtk = true
+    }
+})
+
+-- 目标配置
+target("run")
+    set_kind("binary")
+    
+    -- 添加源文件
+    add_files("main.cpp")
+    add_files("others/src/*.cpp")
+    add_files("others/src/camera/*.cpp")
+    add_files("energy/src/energy/*.cpp")
+    add_files("energy/src/energy/*/*.cpp")
+    add_files("armor/src/armor_finder/*.cpp")
+    add_files("armor/src/armor_finder/*/*.cpp")
+    add_files("armor/src/show_images/*.cpp")
+    
+    -- 添加头文件搜索路径
+    add_includedirs("others/include")
+    add_includedirs("others/include/camera")
+    add_includedirs("others/include/config")
+    add_includedirs("energy/include")
+    add_includedirs("energy/include/energy")
+    add_includedirs("armor/include")
+    add_includedirs("armor/include/armor_finder")
+    add_includedirs("armor/include/armor_finder/classifier")
+    add_includedirs("armor/include/show_images")
+    
+    -- 添加依赖包
+    add_packages("pthread", "eigen", "opencv")
+    
+    -- 添加链接目录
+    add_linkdirs("others")
+    
+    -- 根据平台链接相机 SDK
+    if is_plat("linux") then
+        add_links("MVSDK")
+        print("current platform: Linux")
+    elseif is_plat("windows") then
+        add_links("MVCAMSDK_X64")
+        print("current platform: Windows")
+    elseif is_plat("macosx") then
+        add_links("mvsdk")
+        print("current platform: Mac")
+    else
+        print("Unsupported platform")
+    end
+    
+    -- 设置目标目录
+    set_targetdir("$(builddir)")
+    
+    -- 添加编译选项
+    add_cxxflags("-O3")
+
+-- 自定义任务：create-startup
+task("create-startup")
+    set_category("action")
+    on_run(function ()
+        os.exec("$(projectdir)/tools/create-startup.sh $(projectdir) $(builddir)")
+    end)
+    set_menu {
+        usage = "xmake create-startup",
+        description = "Create startup script",
+        options = {}
+    }
+
+-- 自定义任务：train-cnn
+task("train-cnn")
+    set_category("action")
+    on_run(function ()
+        if os.host() == "linux" then
+            os.exec("gnome-terminal -- bash -c \"$(projectdir)/tools/TrainCNN/backward.py\"")
+        else
+            print("train-cnn only supported on Linux with gnome-terminal")
+        end
+    end)
+    set_menu {
+        usage = "xmake train-cnn",
+        description = "Train CNN model",
+        options = {}
+    }