devbox
菜鸟追梦旅行
这个屌丝很懒，什么也没留下！
关注作者
热门标签
热门文章
当前位置:   article > 正文

Darknet框架简介

作者：菜鸟追梦旅行 | 2024-04-07 18:56:15

赞

踩

darknet

Darknet框架简介

darknet是一个较为轻型的完全基于C与CUDA的开源深度学习框架，其主要特点就是容易安装，没有任何依赖项（OpenCV都可以不用），移植性非常好，支持CPU与GPU两种计算方式。

前言：为什么选择darknet？

  • 相比于TensorFlow来说，darknet并没有那么强大，但这也成了darknet的优势：

  • darknet完全由C语言实现，没有任何依赖项，当然可以使用OpenCV，但只是用其来显示图片、为了更好的可视化；

  • darknet支持CPU（所以没有GPU也不用紧的）与GPU（CUDA/cuDNN，使用GPU当然更块更好了）；

  • 正是因为其较为轻型，没有像TensorFlow那般强大的API，所以给我的感觉就是有另一种味道的灵活性，适合用来研究底层，可以更为方便的从底层对其进行改进与扩展；

  • darknet的实现与caffe的实现存在相似的地方，熟悉了darknet，相信对上手caffe有帮助

linux端

一、下载源码及权重文件

官方darknet源码下载网址：https://github.com/pjreddie/darknet
官方权重下载地址：https://pjreddie.com/media/files/yolov3_tiny.weights
csdn下载地址：yolov3-tiny 权重文件

主要参考：详细步骤！使用yolov3-tiny训练，测试、验证VOC数据集

https://blog.csdn.net/xiao_lxl/article/details/85047103

下载后，首先查看makefile
在这里插入图片描述
可以根据自己实际情况进行更改，用下面命令进行编译

make clean;make all

  • 1

编译完成后生成darknet 可执行文件及动态库。

如下：
在这里插入图片描述
文件夹说明

  • 1.cfg文件夹内是一些模型的架构，每个cfg文件类似与caffe的prototxt文件，通过该文件定义的整个模型的架构

  • 2.data文件夹内放置了一些label文件，如coco9k的类别名等，和一些样例图（该文件夹主要为演示用，或者是直接训练coco等对应数据集时有用，如果要用自己的数据自行训练，该文件夹内的东西都不是我们需要的）

  • 3.src文件夹内全是最底层的框架定义文件，所有层的定义等最基本的函数全部在该文件夹内，可以理解为该文件夹就是框架的源码；

  • 4.examples文件夹是更为高层的一些函数，如检测函数，识别函数等，这些函数直接调用了底层的函数，我们经常使用的就是example中的函数；

  • 5.include文件夹，顾名思义，存放头文件的地方；

  • 6.python文件夹里是使用python对模型的调用方法，基本都在darknet.py中。当然，要实现python的调用，还需要用到darknet的动态库libdarknet.so，这个动态库稍后再介绍；

  • 7.scripts文件夹中是一些脚本，如下载coco数据集，将voc格式的数据集转换为训练所需格式的脚本等。

  • 8.除了license文件，剩下的就是Makefile文件，如下图，在问价开头有一些选项，把你需要使用的选项设为1即可。

二、测试

./darknet detector test cfg/coco.data cfg/yolov3-tiny.cfg weights/yolov3-tiny.weights  data/dog.jpg

  • 1

在这里插入图片描述
出现上述的结果，说明运行正常。

三、训练自己的数据集

待续。。。

windows端

参考：

编译darknet：

https://blog.csdn.net/qq_28663849/article/details/107362445

CPU版本
首先打开makefile中，特别是OPENMP和AVX，速度快

OPENCV=1
AVX=1
OPENMP=1

  • 1
  • 2
  • 3
  • 1、先运行 编译darknet.sln，生成解决方案，release版本，生成darknet.exe。
  • 2、基于上一版本运行 编译yolo_cpp_dll_no_gpu.sln，生成解决方案，release版本，生成yolo_cpp_dll.lib和yolo_cpp_dll.dll。（注意是yolo_cpp_dll）
    3、复制生成的yolo_cpp_dll.lib和yolo_cpp_dll.dll到新工程。
    在这里插入图片描述
    1、动态链接库（均在darknet-master\build\darknet\x64目录下）
    (1)yolo_cpp_dll.lib
    (2)yolo_cpp_dll.dll
    (3)pthreadGC2.dll
    (4)pthreadVC2.dll
    2、OpenCV库（取决于使用debug还是release模式）——或者自己直接在属性管理器中配置。
    (1)opencv_world340d.dll
    (2)opencv_world340.dll
    3、YOLO模型文件（第一个文件在darknet-master\build\darknet\x64\data目录下，第二个文件在darknet-master\build\darknet\x64目录下，第三个文件需要自己下载，下载链接见前一篇文章）
    (1)coco.names
    (2)yolov3.cfg
    (3)yolov3.weights
    4、头文件
    (1)yolo_v2_class.hpp
    头文件包含了动态链接库中具体的类的定义，调用时需要引用，这个文件在darknet-master\include目录下的yolo_v2_class.hpp中，将其复制到记事本保存成.hpp文件即可。

参考：
原文链接：https://blog.csdn.net/stjuliet/article/details/87884976

main.c

#include <iostream>


#define OPENCV
//#define GPU


#include "yolo_v2_class.hpp" //引用动态链接库中的头文件
#include <opencv2/opencv.hpp>
#include "opencv2/highgui/highgui.hpp"

//#pragma comment(lib, "opencv_world340d.lib") //引入OpenCV链接库
//#pragma comment(lib, "yolo_cpp_dll_no_gpu.lib") //引入YOLO动态链接库
#pragma comment(lib, "yolo_cpp_dll.lib") //引入YOLO动态链接库
//以下两段代码来自yolo_console_dll.sln
void draw_boxes(cv::Mat mat_img, std::vector<bbox_t> result_vec, std::vector<std::string> obj_names,
	int current_det_fps , int current_cap_fps )
{
	int const colors[6][3] = { { 1,0,1 },{ 0,0,1 },{ 0,1,1 },{ 0,1,0 },{ 1,1,0 },{ 1,0,0 } };

	for (auto &i : result_vec) 
	{
		cv::Scalar color = obj_id_to_color(i.obj_id);
		cv::rectangle(mat_img, cv::Rect(i.x, i.y, i.w, i.h), color, 2);
		if (obj_names.size() > i.obj_id) 
		{
			std::string obj_name = obj_names[i.obj_id];
			if (i.track_id > 0) obj_name += " - " + std::to_string(i.track_id);
			cv::Size const text_size = getTextSize(obj_name, cv::FONT_HERSHEY_COMPLEX_SMALL, 1.2, 2, 0);
			int const max_width = (text_size.width > i.w + 2) ? text_size.width : (i.w + 2);
			cv::rectangle(mat_img, cv::Point2f(std::max((int)i.x - 1, 0), std::max((int)i.y - 30, 0)),
				cv::Point2f(std::min((int)i.x + max_width, mat_img.cols - 1), std::min((int)i.y, mat_img.rows - 1)),
				color, CV_FILLED, 8, 0);
			putText(mat_img, obj_name, cv::Point2f(i.x, i.y - 10), cv::FONT_HERSHEY_COMPLEX_SMALL, 1.2, cv::Scalar(0, 0, 0), 2);
		}
	}
	if (current_det_fps >= 0) 
	{
		std::string fps_str = "FPS detection: " + std::to_string(current_det_fps) ;
		putText(mat_img, fps_str, cv::Point2f(10, 20), cv::FONT_HERSHEY_COMPLEX_SMALL, 1.2, cv::Scalar(50, 255, 0), 2);
	}
	if (current_cap_fps >= 0)
	{
		std::string fps_str = "FPS capture: " + std::to_string(current_cap_fps);
		putText(mat_img, fps_str, cv::Point2f(10, 50), cv::FONT_HERSHEY_COMPLEX_SMALL, 1.2, cv::Scalar(50, 255, 0), 2);
	}
}

std::vector<std::string> objects_names_from_file(std::string const filename) 
{
	std::ifstream file(filename);
	std::vector<std::string> file_lines;
	if (!file.is_open()) return file_lines;
	for (std::string line; getline(file, line);) file_lines.push_back(line);
	std::cout << "object names loaded \n";
	return file_lines;
}

int main()
{
	std::string names_file = "coco.names";
	std::string cfg_file = "yolov3-tiny.cfg";
	std::string weights_file = "yolov3-tiny.weights";
	Detector detector(cfg_file, weights_file, 0); //初始化检测器
//std::vector<std::string> obj_names = objects_names_from_file(names_file); //调用获得分类对象名称
//或者使用以下四行代码也可实现读入分类对象文件
	std::vector<std::string> obj_names;
	std::ifstream ifs(names_file.c_str());
	std::string line;
	while (getline(ifs, line)) obj_names.push_back(line);
	//测试是否成功读入分类对象文件
	for (size_t i = 0; i < obj_names.size(); i++)
	{
		std::cout << obj_names[i] << std::endl;
	}

	cv::VideoCapture capture;
	capture.open("bsd.mp4");
	if (!capture.isOpened())
	{
		printf("文件打开失败");
	}
	cv::Mat frame;
	float fps = 0;
	while (true)
	{
		int begin, duration;
		begin = clock();
		capture >> frame;
		//if (frame.empty() == NULL)
		//{
		//	printf("Image is empty");
		//	return 0;
		//}


		duration = clock() - begin;
		//printf("读图用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);
		//cv::resize(frame, frame, cv::Size(640,360));
		std::vector<bbox_t> result_vec = detector.detect(frame);
		duration = clock() - begin;
		printf("识别用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);
		fps =(float) 1000.0 / duration;
		draw_boxes(frame, result_vec, obj_names,fps, fps);
		//duration = clock() - begin;
		//printf("画框用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);

		//cv::namedWindow("bsd", 0);
		//cv::imshow("bsd", frame);
		//cv::waitKey(1);
		//duration = clock() - begin;
		//printf("显示用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);
		printf("\n\n\n");

	}
	return 0;
}


  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
  • 37
  • 38
  • 39
  • 40
  • 41
  • 42
  • 43
  • 44
  • 45
  • 46
  • 47
  • 48
  • 49
  • 50
  • 51
  • 52
  • 53
  • 54
  • 55
  • 56
  • 57
  • 58
  • 59
  • 60
  • 61
  • 62
  • 63
  • 64
  • 65
  • 66
  • 67
  • 68
  • 69
  • 70
  • 71
  • 72
  • 73
  • 74
  • 75
  • 76
  • 77
  • 78
  • 79
  • 80
  • 81
  • 82
  • 83
  • 84
  • 85
  • 86
  • 87
  • 88
  • 89
  • 90
  • 91
  • 92
  • 93
  • 94
  • 95
  • 96
  • 97
  • 98
  • 99
  • 100
  • 101
  • 102
  • 103
  • 104
  • 105
  • 106
  • 107
  • 108
  • 109
  • 110
  • 111
  • 112
  • 113
  • 114
  • 115
  • 116
  • 117
  • 118
第二版
/*****************************************************************************************
1、盲区检测

2021.3.17：
*****************************************************************************************/
#define OPENCV/*必须放置在链接的头文件前面*/
//#define GPU

#include <iostream>
#include <stdlib.h> //srand()和rand()函数 
#include<windows.h>
#include <fstream> 

#include "yolo_v2_class.hpp" //引用动态链接库中的头文件

/*opencv相关*/
#include <opencv2/core/core.hpp> 
#include <opencv2/opencv.hpp>
#include "opencv2/highgui/highgui.hpp"
#include <opencv2/imgproc/imgproc.hpp>  
#include <opencv2/objdetect/objdetect.hpp> 
#include <opencv2/ml/ml.hpp>  
#include <opencv2\imgproc\types_c.h>


using namespace cv;
using namespace std;
/****************	     打印相关组件 start !	 ********************************************************************/

#define  debug_show_picture        1 //是否显示部分调试图片，方便调试
/*打印等级，修改后面的宏定义可以改变函数输出打印等级*/
#define ALG_PRTINT_LEVER PRINT_LEVEL_UNLIMIT

#define ALG_PRTINT(...)  SAL_printf(__VA_ARGS__)
#define ALG_PRT(...)     ALG_PRTINT(__FUNCTION__, __LINE__, PRINT_LEVEL_UNLIMIT, __VA_ARGS__)
#define ALG_DBG(...)     ALG_PRTINT(__FUNCTION__, __LINE__, PRINT_LEVEL_DBG,     __VA_ARGS__)
#define ALG_WAR(...)     ALG_PRTINT(__FUNCTION__, __LINE__, PRINT_LEVEL_WRN,     __VA_ARGS__)
#define ALG_ERR(...)     ALG_PRTINT(__FUNCTION__, __LINE__, PRINT_LEVEL_ERR,     __VA_ARGS__)

/***********************************************************************************************
* @enum     HAT_SAL_PRT_LEVEL_E
* @brief    打印输出的等级
***************************************************************************************************/
typedef enum _PRT_LEVEL_E_
{
	PRINT_LEVEL_ERR = 0,        /*错误打印信息*/
	PRINT_LEVEL_WRN = 1,        /*警告打印信息*/
	PRINT_LEVEL_DBG = 2,         /*调试打印信息*/
	PRINT_LEVEL_UNLIMIT = 3,    /*无限制打印信息*/
	PRINT_LEVEL_NOPRT = 4,      /*没有打印信息*/
} PRT_LEVEL_E;

/*******************************************************************************
Function:	SAL_printf
Description: 该函数能够通过设置的打印等级ALG_PRTINT_LEVER，来控制是否输出相关语句
Input:
Output:
Return:
0:			Successful
ohters:		Failed
*******************************************************************************/
void SAL_printf(const char *pFun, UINT line, PRT_LEVEL_E levelParam, const char *fmt, ...)
{
	static INT8 g_printfInfo[4][16] = { "ERR", "WAR", "DBG", "INF" };
	va_list p;
	if (ALG_PRTINT_LEVER == PRINT_LEVEL_NOPRT || levelParam == PRINT_LEVEL_NOPRT)
	{
		return;
	}
	if (levelParam <= ALG_PRTINT_LEVER)
	{
		va_start(p, fmt);
		printf("[DIYUN_ALG][%s][%s][%4d] ", g_printfInfo[levelParam], pFun, line);
		vprintf(fmt, p);
		va_end(p);
	}
}
/******************************  打印相关组件 end   *************************************************************************/









//#pragma comment(lib, "opencv_world340d.lib") //引入OpenCV链接库
//#pragma comment(lib, "yolo_cpp_dll_no_gpu.lib") //引入YOLO动态链接库
#pragma comment(lib, "yolo_cpp_dll.lib") //引入YOLO动态链接库

#ifdef max
#undef max
#endif

#ifdef min
#undef min
#endif

//以下两段代码来自yolo_console_dll.sln
void draw_boxes(Mat mat_img, vector<bbox_t> result_vec, vector<string> obj_names,
	int current_det_fps , int current_cap_fps )
{
	int const colors[6][3] = { { 1,0,1 },{ 0,0,1 },{ 0,1,1 },{ 0,1,0 },{ 1,1,0 },{ 1,0,0 } };

	for (auto &i : result_vec) 
	{
		Scalar color = obj_id_to_color(i.obj_id);
		rectangle(mat_img, Rect(i.x, i.y, i.w, i.h), color, 2);
		if (obj_names.size() > i.obj_id) 
		{
			string obj_name = obj_names[i.obj_id];
			if (i.track_id > 0) obj_name += " - " + to_string(i.track_id);
			Size const text_size = getTextSize(obj_name, FONT_HERSHEY_COMPLEX_SMALL, 1.2, 2, 0);
			int const max_width = (text_size.width > i.w + 2) ? text_size.width : (i.w + 2);
			rectangle(mat_img, Point2f(max((int)i.x - 1, 0), max((int)i.y - 30, 0)),
				Point2f(min((int)i.x + max_width, mat_img.cols - 1), min((int)i.y, mat_img.rows - 1)),
				color, CV_FILLED, 8, 0);
			putText(mat_img, obj_name, Point2f(i.x, i.y - 10), FONT_HERSHEY_COMPLEX_SMALL, 1.2, Scalar(0, 0, 0), 2);
		}
	}
	if (current_det_fps >= 0) 
	{
		string fps_str = "FPS detection: " + to_string(current_det_fps) ;
		putText(mat_img, fps_str, Point2f(10, 20), FONT_HERSHEY_COMPLEX_SMALL, 1.2, Scalar(50, 255, 0), 2);
	}
	if (current_cap_fps >= 0)
	{
		string fps_str = "FPS capture: " + to_string(current_cap_fps);
		putText(mat_img, fps_str, Point2f(10, 50), FONT_HERSHEY_COMPLEX_SMALL, 1.2, Scalar(50, 255, 0), 2);
	}
}

vector<string> objects_names_from_file(string const filename) 
{
	ifstream file(filename);
	vector<string> file_lines;
	if (!file.is_open()) return file_lines;
	for (string line; getline(file, line);) file_lines.push_back(line);
	cout << "object names loaded \n";
	return file_lines;
}

//点(浮点型) 
typedef struct _POINT_F_
{
	float x;
	float y;
}POINT_F;

//多边形(浮点型)
typedef struct _VCA_POLYGON_F_
{
	unsigned int  vertex_num;                  //顶点数
	POINT_F   point[5];   //顶点
}VCA_POLYGON_F;

//前后视图片切换索引
typedef enum _BSD_WARNNING_TYPE_
{
	BSD_WARNNING_INVALID = 0,              // 无效值
	BSD_WARNNING_WARN = 1,              // BSD提醒（超车）
	BSD_WARNNING_ALARM = 2,              // BSD预警（无超车要求）
	BSD_WARNNING_ICW = 3,              // 摄像头失效报警
	BSD_EVENT_MOD = 4,              // 移动障碍物检测
	BSD_EVENT_MOD_ALL = 5,              // 移动障碍物检测(全景)
	BSD_EVENT_DOW = 6,              // 开门预警检测
	BSD_WARNNING_END
} BSD_WARNNING_TYPE;

// BSDC报警区域枚举
typedef enum _VCA_REAR_VIEW_STATE_
{
	VCA_MOD_INVALID = 0,           // 无障碍物（无效）
	VCA_MOD_OBS = 1,           // 近处有障碍物
	VCA_MOD_FAR = 2,           // 远处有障碍物
	VCA_MOD_LEFT = 3,           // 左侧有障碍物
	VCA_MOD_RIGHT = 4,           // 右侧有障碍物
	VCA_MOD_LEFT_LEVEL_TWO = 5,    // 左侧有障碍物二级报警
	VCA_MOD_RIGHT_LEVEL_TWO = 6,   // 右侧有障碍物二级报警
	VCA_MOD_LEVEL1 = 11,           // 一级区域——最危险
	VCA_MOD_LEVEL2 = 12,           // 二级区域——中等危险
	VCA_MOD_LEVEL3 = 13,           // 三级区域——较安全

	VCA_MOD_END
} VCA_REAR_VIEW_STATE;

/* VCA_BSD配置参数 */
typedef struct _BSD_REGION_PARAM_SINGLE_
{
	BSD_WARNNING_TYPE     	event_type;/*报警类型*/
	VCA_REAR_VIEW_STATE   	view_state;			/* 报警区域级别 */
	VCA_POLYGON_F         	poly;      /*报警框顶点*/
	short                 	sensity;
	char                  	reserved[2];
} BSD_REGION_PARAM_SINGLE;

/* BSD配置参数 */
typedef struct _BSD_PARAM_ALL_
{
	int                     region_num;
	BSD_REGION_PARAM_SINGLE	region[4];
	int			time_type;/*0 1 2*/
	char					mirror_flag;		/* 图像和结果是否需要镜像，1为镜像，0不镜像 */
} VCA_BSD_PARAM_ALL;

// 侧视、车尾的报警区域配置参数
typedef struct _VCA_ADAS_BSDC_MOD_INPUT_PARAM_
{
	int                   whole_w;
	int                   whole_h;
	int                   poly_valid[3];  // 报警区域是否有效标志位
	VCA_POLYGON_F         bsd_poly[3];    // BSD报警区域坐标
	VCA_REAR_VIEW_STATE   view_state[3];  // 报警区域级别
} VCA_ADAS_BSDC_MOD_INPUT_PARAM;

/*全局变量，用于bsd参数配置传递等*/
VCA_ADAS_BSDC_MOD_INPUT_PARAM gBsd_init_para = { 0 };

/*******************************************************************************
函数名称:	BsdInitParam
描    述: 该函数初始化BSD报警参数配置，主要是报警框的设置
输入:输入宽高
输出:
返回值:
		0:			Successful
		ohters:		Failed
*******************************************************************************/
int BsdInitParam(int  width,int height)
{
	bool OK = 0;/*返回值初始化*/
	int ori_w = width;
	int ori_h = height;

	VCA_ADAS_BSDC_MOD_INPUT_PARAM *pBsd_init_para = NULL;
	pBsd_init_para = &gBsd_init_para;
#if 0
	/*左视*/
	//一级区域(一般为最小、最靠近本车的区域)
	pBsd_init_para->poly_valid[0] = 1;                     //该区域有效
	pBsd_init_para->view_state[0] = VCA_MOD_LEVEL3;          //一级区域
	pBsd_init_para->bsd_poly[0].vertex_num = 5;
	pBsd_init_para->bsd_poly[0].point[0].x = (float)1100 / ori_w;
	pBsd_init_para->bsd_poly[0].point[1].x = (float)1255 / ori_w;
	pBsd_init_para->bsd_poly[0].point[2].x = (float)1083 / ori_w;
	pBsd_init_para->bsd_poly[0].point[3].x = (float)643 / ori_w;
	pBsd_init_para->bsd_poly[0].point[4].x = (float)643 / ori_w;

	pBsd_init_para->bsd_poly[0].point[0].y = (float)238 / ori_h;
	pBsd_init_para->bsd_poly[0].point[1].y = (float)218 / ori_h;
	pBsd_init_para->bsd_poly[0].point[2].y = (float)713 / ori_h;
	pBsd_init_para->bsd_poly[0].point[3].y = (float)716 / ori_h;
	pBsd_init_para->bsd_poly[0].point[4].y = (float)626 / ori_h;

	//二级区域 
	pBsd_init_para->poly_valid[1] = 1;                     //该区域有效 
	pBsd_init_para->view_state[1] = VCA_MOD_LEVEL2;          //二级区域
	pBsd_init_para->bsd_poly[0].vertex_num = 5;
	pBsd_init_para->bsd_poly[0].point[0].x = (float)1040 / ori_w;
	pBsd_init_para->bsd_poly[0].point[1].x = (float)1265 / ori_w;
	pBsd_init_para->bsd_poly[0].point[2].x = (float)1083 / ori_w;
	pBsd_init_para->bsd_poly[0].point[3].x = (float)413 / ori_w;
	pBsd_init_para->bsd_poly[0].point[4].x = (float)413 / ori_w;

	pBsd_init_para->bsd_poly[0].point[0].y = (float)218 / ori_h;
	pBsd_init_para->bsd_poly[0].point[1].y = (float)198 / ori_h;
	pBsd_init_para->bsd_poly[0].point[2].y = (float)716 / ori_h;
	pBsd_init_para->bsd_poly[0].point[3].y = (float)716 / ori_h;
	pBsd_init_para->bsd_poly[0].point[4].y = (float)616 / ori_h;

	//三级区域
	pBsd_init_para->poly_valid[2] = 1; 					 //该区域有效
	pBsd_init_para->view_state[2] = VCA_MOD_LEVEL1;			 //三级区域
	pBsd_init_para->bsd_poly[2].vertex_num = 5;
	pBsd_init_para->bsd_poly[2].point[0].x = (float)960 / ori_w;	//	925
	pBsd_init_para->bsd_poly[2].point[1].x = (float)1265 / ori_w; // 
	pBsd_init_para->bsd_poly[2].point[2].x = (float)1083 / ori_w;
	pBsd_init_para->bsd_poly[2].point[3].x = (float)183 / ori_w;
	pBsd_init_para->bsd_poly[2].point[4].x = (float)183 / ori_w;

	pBsd_init_para->bsd_poly[2].point[0].y = (float)198 / ori_h;
	pBsd_init_para->bsd_poly[2].point[1].y = (float)178 / ori_h;
	pBsd_init_para->bsd_poly[2].point[2].y = (float)716 / ori_h;
	pBsd_init_para->bsd_poly[2].point[3].y = (float)716 / ori_h;
	pBsd_init_para->bsd_poly[2].point[4].y = (float)616 / ori_h;

#endif 

# if 1
//	//右视
	pBsd_init_para->poly_valid[0] = 1;                     //该区域有效
	pBsd_init_para->view_state[0] = VCA_MOD_LEVEL3;          //一级区域，最大框
	pBsd_init_para->bsd_poly[0].vertex_num = 5;
	pBsd_init_para->bsd_poly[0].point[0].x = (float)431/ori_w;
	pBsd_init_para->bsd_poly[0].point[0].y = (float)182/ori_h;
	pBsd_init_para->bsd_poly[0].point[1].x = (float)52/ori_w;
	pBsd_init_para->bsd_poly[0].point[1].y = (float)163/ori_h;
	pBsd_init_para->bsd_poly[0].point[2].x = (float)53/ori_w;
	pBsd_init_para->bsd_poly[0].point[2].y =(float) 720/ori_h;
	pBsd_init_para->bsd_poly[0].point[3].x =(float) 704/ori_w;
	pBsd_init_para->bsd_poly[0].point[3].y = (float)712/ori_h;
	pBsd_init_para->bsd_poly[0].point[4].x = (float)579/ori_w;
	pBsd_init_para->bsd_poly[0].point[4].y = (float)480/ori_h;

	//二级区域 
	pBsd_init_para->poly_valid[1] = 1;                     //该区域有效 
	pBsd_init_para->view_state[1] = VCA_MOD_LEVEL2;          //二级区域
	pBsd_init_para->bsd_poly[0].vertex_num = 5;
	pBsd_init_para->bsd_poly[1].point[0].x = (float)298/ori_w;
	pBsd_init_para->bsd_poly[1].point[0].y = (float)187/ori_h;
	pBsd_init_para->bsd_poly[1].point[1].x = (float)60/ori_w;
	pBsd_init_para->bsd_poly[1].point[1].y = (float)172/ori_h;
	pBsd_init_para->bsd_poly[1].point[2].x = (float)66/ori_w;
	pBsd_init_para->bsd_poly[1].point[2].y = (float)717/ori_h;
	pBsd_init_para->bsd_poly[1].point[3].x = (float)481/ori_w;
	pBsd_init_para->bsd_poly[1].point[3].y = (float)709/ori_h;
	pBsd_init_para->bsd_poly[1].point[4].x = (float)400/ori_w;
	pBsd_init_para->bsd_poly[1].point[4].y = (float)495/ori_h;

	//三级区域
	pBsd_init_para->poly_valid[2] = 1; 					 //该区域有效
	pBsd_init_para->view_state[2] = VCA_MOD_LEVEL1;			 //三级区域
	pBsd_init_para->bsd_poly[2].vertex_num = 5;
	pBsd_init_para->bsd_poly[2].point[0].x = (float)183/ori_w;
	pBsd_init_para->bsd_poly[2].point[0].y = (float)172/ori_h;
	pBsd_init_para->bsd_poly[2].point[1].x = (float)75/ori_w;
	pBsd_init_para->bsd_poly[2].point[1].y = (float)171/ori_h;
	pBsd_init_para->bsd_poly[2].point[2].x = (float)46/ori_w;
	pBsd_init_para->bsd_poly[2].point[2].y = (float)705/ori_h;
	pBsd_init_para->bsd_poly[2].point[3].x = (float)234/ori_w;
	pBsd_init_para->bsd_poly[2].point[3].y = (float)706/ori_h;
	pBsd_init_para->bsd_poly[2].point[4].x = (float)218/ori_w;
	pBsd_init_para->bsd_poly[2].point[4].y = (float)558/ori_h;

#endif




	return OK;
}

/*******************************************************************************
函数名称:	Bsd_draw_jeiguo
描    述: 该函数画图
输入:输入宽高,以及报警框个数
输出:
返回值: 图像
0:			Successful
ohters:		Failed
*******************************************************************************/
Mat Bsd_draw_jeiguo(Mat src, int  width, int height,int lever)
{
	bool OK = 0;/*返回值初始化*/
	int ori_w = width;
	int ori_h = height;
	Mat dst;

	VCA_ADAS_BSDC_MOD_INPUT_PARAM *pBsd_init_para = NULL;
	pBsd_init_para = &gBsd_init_para;
	Point p1,p2;
	/*绘制报警框*/
	for (int i = 0; i < lever; i++)/*三个等级*/
	{
		for (int j = 0; j < 5; j++)/*每个报警框由5条线段组成*/
		{
			p1.x = (int )( pBsd_init_para->bsd_poly[i].point[j].x*ori_w );
			p1.y = (int )( pBsd_init_para->bsd_poly[i].point[j].y*ori_h );
			if (j == 4)
			{
				p2.x = (int )(pBsd_init_para->bsd_poly[i].point[0].x*ori_w);
				p2.y = (int)( pBsd_init_para->bsd_poly[i].point[0].y*ori_h);
			}
			else
			{
				p2.x = (int)( pBsd_init_para->bsd_poly[i].point[j + 1].x*ori_w);
				p2.y = (int)( pBsd_init_para->bsd_poly[i].point[j + 1].y*ori_h);
			}
			/*绘制直线*/
			line(src, p1, p2, Scalar(0, 0, 255), 2);
		}
	}
	return src;
}

/*******************************************************************************
函数名称:	Bsd_AlarmInfo
描    述: 判断报警等级
输入:输入报警框,
输出:报警结果
返回值: 图像
0:			Successful
ohters:		Failed
*******************************************************************************/
Mat Bsd_AlarmInfo(Mat src, int  width, int height, int lever)
{
	bool OK = 0;/*返回值初始化*/
	int ori_w = width;
	int ori_h = height;
	Mat dst;

	VCA_ADAS_BSDC_MOD_INPUT_PARAM *pBsd_init_para = NULL;
	pBsd_init_para = &gBsd_init_para;
	Point p1, p2;
	/*绘制报警框*/
	for (int i = 0; i < lever; i++)/*三个等级*/
	{
		for (int j = 0; j < 5; j++)/*每个报警框由5条线段组成*/
		{
			p1.x = (int)(pBsd_init_para->bsd_poly[i].point[j].x*ori_w);
			p1.y = (int)(pBsd_init_para->bsd_poly[i].point[j].y*ori_h);
			if (j == 4)
			{
				p2.x = (int)(pBsd_init_para->bsd_poly[i].point[0].x*ori_w);
				p2.y = (int)(pBsd_init_para->bsd_poly[i].point[0].y*ori_h);
			}
			else
			{
				p2.x = (int)(pBsd_init_para->bsd_poly[i].point[j + 1].x*ori_w);
				p2.y = (int)(pBsd_init_para->bsd_poly[i].point[j + 1].y*ori_h);
			}
			/*绘制直线*/
			line(src, p1, p2, Scalar(0, 0, 255), 2);
		}
	}
	return src;
}







int main()
{
	bool OK = 0;/*返回值初始化*/
	int ret = TRUE;
	int BSD_IMAGE_W = 1280;/*宽度*/
	int BSD_IMAGE_H = 1280;/*高度*/
	int lever = 1;/*只显示一个报警框*/
	//string names_file = "coco.names";
	//string cfg_file = "yolov3-tiny.cfg";
	//string weights_file = "yolov3-tiny.weights";
	/*test*/
	string names_file = "coco.names";
	string cfg_file = "yolov3-tiny.cfg";
	string weights_file = "yolov3-tiny.weights";

	//string names_file = "coco.names";
	//string cfg_file = "yolov4-tiny.cfg";
	//string weights_file = "yolov4-tiny.weights";



	Detector detector(cfg_file, weights_file, 0); //初始化检测器
	vector<string> obj_names;
	ifstream ifs(names_file.c_str());
	string line;
	while (getline(ifs, line)) obj_names.push_back(line);
	//测试是否成功读入分类对象文件
	for (size_t i = 0; i < obj_names.size(); i++)
	{
		cout << obj_names[i] << endl;
	}
	/*初始化BSD参数*/
	ret = BsdInitParam(BSD_IMAGE_W, BSD_IMAGE_H);
	if (ret != OK)
	{
		ALG_ERR("配置BSD参数有问题\n");
	}

	VideoCapture capture;
	//capture.open("yejianbsd.mp4");
	capture.open("bsd.mp4");
	if (!capture.isOpened())
	{
		printf("文件打开失败");
	}
	Mat frame;
	float fps = 0;
	while (true)
	{
		int begin, duration;
		begin = clock();
		capture >> frame;

		if (frame.data == NULL)
		{
			printf("Image is empty");
			return 0;
		}
		
		duration = clock() - begin;
		//printf("读图用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);
		//resize(frame, frame, Size(640,360));
		vector<bbox_t> result_vec = detector.detect(frame);
		duration = clock() - begin;
		printf("识别用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);
		fps =(float) 1000.0 / duration;
		draw_boxes(frame, result_vec, obj_names,fps, fps);
		//duration = clock() - begin;
		//printf("画框用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);
		frame = Bsd_draw_jeiguo(frame, BSD_IMAGE_W, BSD_IMAGE_H, lever);
		namedWindow("bsd", 0);
		imshow("bsd", frame);
		waitKey(1);
		//duration = clock() - begin;
		//printf("显示用时约: %d毫秒 \n", duration * 1000 / CLOCKS_PER_SEC);
		printf("\n\n\n");
	}

	return 0;
}


  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
  • 37
  • 38
  • 39
  • 40
  • 41
  • 42
  • 43
  • 44
  • 45
  • 46
  • 47
  • 48
  • 49
  • 50
  • 51
  • 52
  • 53
  • 54
  • 55
  • 56
  • 57
  • 58
  • 59
  • 60
  • 61
  • 62
  • 63
  • 64
  • 65
  • 66
  • 67
  • 68
  • 69
  • 70
  • 71
  • 72
  • 73
  • 74
  • 75
  • 76
  • 77
  • 78
  • 79
  • 80
  • 81
  • 82
  • 83
  • 84
  • 85
  • 86
  • 87
  • 88
  • 89
  • 90
  • 91
  • 92
  • 93
  • 94
  • 95
  • 96
  • 97
  • 98
  • 99
  • 100
  • 101
  • 102
  • 103
  • 104
  • 105
  • 106
  • 107
  • 108
  • 109
  • 110
  • 111
  • 112
  • 113
  • 114
  • 115
  • 116
  • 117
  • 118
  • 119
  • 120
  • 121
  • 122
  • 123
  • 124
  • 125
  • 126
  • 127
  • 128
  • 129
  • 130
  • 131
  • 132
  • 133
  • 134
  • 135
  • 136
  • 137
  • 138
  • 139
  • 140
  • 141
  • 142
  • 143
  • 144
  • 145
  • 146
  • 147
  • 148
  • 149
  • 150
  • 151
  • 152
  • 153
  • 154
  • 155
  • 156
  • 157
  • 158
  • 159
  • 160
  • 161
  • 162
  • 163
  • 164
  • 165
  • 166
  • 167
  • 168
  • 169
  • 170
  • 171
  • 172
  • 173
  • 174
  • 175
  • 176
  • 177
  • 178
  • 179
  • 180
  • 181
  • 182
  • 183
  • 184
  • 185
  • 186
  • 187
  • 188
  • 189
  • 190
  • 191
  • 192
  • 193
  • 194
  • 195
  • 196
  • 197
  • 198
  • 199
  • 200
  • 201
  • 202
  • 203
  • 204
  • 205
  • 206
  • 207
  • 208
  • 209
  • 210
  • 211
  • 212
  • 213
  • 214
  • 215
  • 216
  • 217
  • 218
  • 219
  • 220
  • 221
  • 222
  • 223
  • 224
  • 225
  • 226
  • 227
  • 228
  • 229
  • 230
  • 231
  • 232
  • 233
  • 234
  • 235
  • 236
  • 237
  • 238
  • 239
  • 240
  • 241
  • 242
  • 243
  • 244
  • 245
  • 246
  • 247
  • 248
  • 249
  • 250
  • 251
  • 252
  • 253
  • 254
  • 255
  • 256
  • 257
  • 258
  • 259
  • 260
  • 261
  • 262
  • 263
  • 264
  • 265
  • 266
  • 267
  • 268
  • 269
  • 270
  • 271
  • 272
  • 273
  • 274
  • 275
  • 276
  • 277
  • 278
  • 279
  • 280
  • 281
  • 282
  • 283
  • 284
  • 285
  • 286
  • 287
  • 288
  • 289
  • 290
  • 291
  • 292
  • 293
  • 294
  • 295
  • 296
  • 297
  • 298
  • 299
  • 300
  • 301
  • 302
  • 303
  • 304
  • 305
  • 306
  • 307
  • 308
  • 309
  • 310
  • 311
  • 312
  • 313
  • 314
  • 315
  • 316
  • 317
  • 318
  • 319
  • 320
  • 321
  • 322
  • 323
  • 324
  • 325
  • 326
  • 327
  • 328
  • 329
  • 330
  • 331
  • 332
  • 333
  • 334
  • 335
  • 336
  • 337
  • 338
  • 339
  • 340
  • 341
  • 342
  • 343
  • 344
  • 345
  • 346
  • 347
  • 348
  • 349
  • 350
  • 351
  • 352
  • 353
  • 354
  • 355
  • 356
  • 357
  • 358
  • 359
  • 360
  • 361
  • 362
  • 363
  • 364
  • 365
  • 366
  • 367
  • 368
  • 369
  • 370
  • 371
  • 372
  • 373
  • 374
  • 375
  • 376
  • 377
  • 378
  • 379
  • 380
  • 381
  • 382
  • 383
  • 384
  • 385
  • 386
  • 387
  • 388
  • 389
  • 390
  • 391
  • 392
  • 393
  • 394
  • 395
  • 396
  • 397
  • 398
  • 399
  • 400
  • 401
  • 402
  • 403
  • 404
  • 405
  • 406
  • 407
  • 408
  • 409
  • 410
  • 411
  • 412
  • 413
  • 414
  • 415
  • 416
  • 417
  • 418
  • 419
  • 420
  • 421
  • 422
  • 423
  • 424
  • 425
  • 426
  • 427
  • 428
  • 429
  • 430
  • 431
  • 432
  • 433
  • 434
  • 435
  • 436
  • 437
  • 438
  • 439
  • 440
  • 441
  • 442
  • 443
  • 444
  • 445
  • 446
  • 447
  • 448
  • 449
  • 450
  • 451
  • 452
  • 453
  • 454
  • 455
  • 456
  • 457
  • 458
  • 459
  • 460
  • 461
  • 462
  • 463
  • 464
  • 465
  • 466
  • 467
  • 468
  • 469
  • 470
  • 471
  • 472
  • 473
  • 474
  • 475
  • 476
  • 477
  • 478
  • 479
  • 480
  • 481
  • 482
  • 483
  • 484
  • 485
  • 486
  • 487
  • 488
  • 489
  • 490
  • 491
  • 492
  • 493
  • 494
  • 495
  • 496
  • 497
  • 498
  • 499
  • 500
  • 501
  • 502
  • 503
  • 504
  • 505
  • 506
  • 507
  • 508
  • 509
  • 510
  • 511
  • 512
  • 513
  • 514
  • 515

纯cpu效果图：
在这里插入图片描述

如果只跑cpu，需要把darknet.vcxproj文件中cuda和cudnn路径去掉，不然打不开，
如下：

<?xml version="1.0" encoding="utf-8"?>
<Project DefaultTargets="Build" ToolsVersion="14.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
  <ItemGroup Label="ProjectConfigurations">
    <ProjectConfiguration Include="Debug|Win32">
      <Configuration>Debug</Configuration>
      <Platform>Win32</Platform>
    </ProjectConfiguration>
    <ProjectConfiguration Include="Debug|x64">
      <Configuration>Debug</Configuration>
      <Platform>x64</Platform>
    </ProjectConfiguration>
    <ProjectConfiguration Include="Release|Win32">
      <Configuration>Release</Configuration>
      <Platform>Win32</Platform>
    </ProjectConfiguration>
    <ProjectConfiguration Include="Release|x64">
      <Configuration>Release</Configuration>
      <Platform>x64</Platform>
    </ProjectConfiguration>
  </ItemGroup>
  <PropertyGroup Label="Globals">
    <ProjectGuid>{4CF5694F-12A5-4012-8D94-9A0915E9FEB5}</ProjectGuid>
    <RootNamespace>darknet</RootNamespace>
    <WindowsTargetPlatformVersion>8.1</WindowsTargetPlatformVersion>
  </PropertyGroup>
  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
    <ConfigurationType>Application</ConfigurationType>
    <UseDebugLibraries>true</UseDebugLibraries>
    <PlatformToolset>v140</PlatformToolset>
    <CharacterSet>MultiByte</CharacterSet>
  </PropertyGroup>
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration">
    <ConfigurationType>Application</ConfigurationType>
    <UseDebugLibraries>true</UseDebugLibraries>
    <PlatformToolset>v140</PlatformToolset>
    <CharacterSet>MultiByte</CharacterSet>
  </PropertyGroup>
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration">
    <ConfigurationType>Application</ConfigurationType>
    <UseDebugLibraries>false</UseDebugLibraries>
    <PlatformToolset>v140</PlatformToolset>
    <WholeProgramOptimization>true</WholeProgramOptimization>
    <CharacterSet>MultiByte</CharacterSet>
  </PropertyGroup>
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration">
    <ConfigurationType>Application</ConfigurationType>
    <UseDebugLibraries>false</UseDebugLibraries>
    <PlatformToolset>v140</PlatformToolset>
    <WholeProgramOptimization>true</WholeProgramOptimization>
    <CharacterSet>MultiByte</CharacterSet>
  </PropertyGroup>
  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
  <ImportGroup Label="ExtensionSettings" />
  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
  </ImportGroup>
  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="PropertySheets">
    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
  </ImportGroup>
  <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
  </ImportGroup>
  <ImportGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="PropertySheets">
    <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
  </ImportGroup>
  <PropertyGroup Label="UserMacros" />
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
    <OutDir>$(SolutionDir)$(Platform)\</OutDir>
  </PropertyGroup>
  <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
    <OutDir>$(SolutionDir)$(Platform)\</OutDir>
  </PropertyGroup>
  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
    <ClCompile>
      <WarningLevel>Level3</WarningLevel>
      <Optimization>Disabled</Optimization>
      <SDLCheck>true</SDLCheck>
    </ClCompile>
    <Link>
      <GenerateDebugInformation>true</GenerateDebugInformation>
    </Link>
  </ItemDefinitionGroup>
  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
    <ClCompile>
      <WarningLevel>Level3</WarningLevel>
      <Optimization>Disabled</Optimization>
      <SDLCheck>true</SDLCheck>
      <AdditionalIncludeDirectories>$(OPENCV_DIR)\include;C:\opencv_3.0\opencv\build\include;..\..\include;..\..\3rdparty\stb\include;..\..\3rdparty\pthreads\include;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
      <PreprocessorDefinitions>_MBCS;LIB_EXPORTS;_TIMESPEC_DEFINED;_SCL_SECURE_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions>
      <UndefinePreprocessorDefinitions>CUDNN</UndefinePreprocessorDefinitions>
      <MultiProcessorCompilation>true</MultiProcessorCompilation>
      <ForcedIncludeFiles>stdlib.h;crtdbg.h;%(ForcedIncludeFiles)</ForcedIncludeFiles>
    </ClCompile>
    <Link>
      <GenerateDebugInformation>true</GenerateDebugInformation>
      <AdditionalLibraryDirectories>$(OPENCV_DIR)\x64\vc15\lib;$(OPENCV_DIR)\x64\vc14\lib;C:\opencv_3.0\opencv\build\x64\vc14\lib;..\..\3rdparty\pthreads\lib;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
      <OutputFile>$(OutDir)\$(TargetName)$(TargetExt)</OutputFile>
      <AdditionalDependencies>pthreadVC2.lib;%(AdditionalDependencies)</AdditionalDependencies>
      <AssemblyDebug>true</AssemblyDebug>
    </Link>
    <CudaCompile>
      <CodeGeneration>compute_30,sm_30;compute_75,sm_75</CodeGeneration>
      <TargetMachinePlatform>64</TargetMachinePlatform>
    </CudaCompile>
  </ItemDefinitionGroup>
  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
    <ClCompile>
      <WarningLevel>Level3</WarningLevel>
      <Optimization>MaxSpeed</Optimization>
      <FunctionLevelLinking>true</FunctionLevelLinking>
      <IntrinsicFunctions>true</IntrinsicFunctions>
      <SDLCheck>true</SDLCheck>
      <AdditionalIncludeDirectories>C:\opencv_2.4.9\opencv\build\include;..\..\include;..\..\3rdparty\include;%(AdditionalIncludeDirectories);$(CudaToolkitIncludeDir);$(cudnn)\include</AdditionalIncludeDirectories>
      <PreprocessorDefinitions>LIB_EXPORTS;_TIMESPEC_DEFINED;_CRT_SECURE_NO_WARNINGS;GPU;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions>
      <MultiProcessorCompilation>true</MultiProcessorCompilation>
    </ClCompile>
    <Link>
      <GenerateDebugInformation>true</GenerateDebugInformation>
      <EnableCOMDATFolding>true</EnableCOMDATFolding>
      <OptimizeReferences>true</OptimizeReferences>
      <AdditionalLibraryDirectories>C:\opencv_2.4.9\opencv\build\x86\vc14\lib;C:\opencv_2.4.9\opencv\build\x86\vc12\lib;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
      <AdditionalDependencies>..\..\3rdparty\lib\x86\pthreadVC2.lib;%(AdditionalDependencies)</AdditionalDependencies>
    </Link>
  </ItemDefinitionGroup>
  <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
    <ClCompile>
      <WarningLevel>Level3</WarningLevel>
      <Optimization>MaxSpeed</Optimization>
      <FunctionLevelLinking>true</FunctionLevelLinking>
      <IntrinsicFunctions>true</IntrinsicFunctions>
      <SDLCheck>true</SDLCheck>
      <AdditionalIncludeDirectories>$(OPENCV_DIR)\include;C:\opencv_3.0\opencv\build\include;..\..\include;..\..\3rdparty\stb\include;..\..\3rdparty\pthreads\include;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
      <PreprocessorDefinitions>LIB_EXPORTS;_TIMESPEC_DEFINED;_SCL_SECURE_NO_WARNINGS;_CRT_SECURE_NO_WARNINGS;WIN32;NDEBUG;_CONSOLE;_LIB;%(PreprocessorDefinitions)</PreprocessorDefinitions>
      <CLanguageStandard>c11</CLanguageStandard>
      <CppLanguageStandard>c++1y</CppLanguageStandard>
      <PrecompiledHeaderCompileAs>CompileAsCpp</PrecompiledHeaderCompileAs>
      <CompileAs>Default</CompileAs>
      <UndefinePreprocessorDefinitions>NDEBUG</UndefinePreprocessorDefinitions>
      <MultiProcessorCompilation>true</MultiProcessorCompilation>
      <AdditionalUsingDirectories>
      </AdditionalUsingDirectories>
    </ClCompile>
    <Link>
      <GenerateDebugInformation>true</GenerateDebugInformation>
      <EnableCOMDATFolding>true</EnableCOMDATFolding>
      <OptimizeReferences>true</OptimizeReferences>
      <AdditionalLibraryDirectories>$(OPENCV_DIR)\x64\vc15\lib;$(OPENCV_DIR)\x64\vc14\lib;C:\opencv_3.0\opencv\build\x64\vc14\lib;..\..\3rdparty\pthreads\lib;%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
      <AdditionalDependencies>pthreadVC2.lib;%(AdditionalDependencies)</AdditionalDependencies>
      <OutputFile>$(OutDir)\$(TargetName)$(TargetExt)</OutputFile>
    </Link>
    <CudaCompile>
      <TargetMachinePlatform>64</TargetMachinePlatform>
      <CodeGeneration>compute_35,sm_35;compute_86,sm_86</CodeGeneration>
    </CudaCompile>
  </ItemDefinitionGroup>
  <ItemGroup>
    <CudaCompile Include="..\..\src\activation_kernels.cu" />
    <CudaCompile Include="..\..\src\avgpool_layer_kernels.cu" />
    <CudaCompile Include="..\..\src\blas_kernels.cu" />
    <CudaCompile Include="..\..\src\col2im_kernels.cu" />
    <CudaCompile Include="..\..\src\convolutional_kernels.cu" />
    <CudaCompile Include="..\..\src\crop_layer_kernels.cu" />
    <CudaCompile Include="..\..\src\deconvolutional_kernels.cu" />
    <CudaCompile Include="..\..\src\dropout_layer_kernels.cu" />
    <CudaCompile Include="..\..\src\im2col_kernels.cu" />
    <CudaCompile Include="..\..\src\maxpool_layer_kernels.cu" />
    <CudaCompile Include="..\..\src\network_kernels.cu" />
  </ItemGroup>
  <ItemGroup>
    <ClCompile Include="..\..\src\activations.c" />
    <ClCompile Include="..\..\src\activation_layer.c" />
    <ClCompile Include="..\..\src\art.c" />
    <ClCompile Include="..\..\src\avgpool_layer.c" />
    <ClCompile Include="..\..\src\batchnorm_layer.c" />
    <ClCompile Include="..\..\src\blas.c" />
    <ClCompile Include="..\..\src\box.c" />
    <ClCompile Include="..\..\src\captcha.c" />
    <ClCompile Include="..\..\src\cifar.c" />
    <ClCompile Include="..\..\src\classifier.c" />
    <ClCompile Include="..\..\src\coco.c" />
    <ClCompile Include="..\..\src\col2im.c" />
    <ClCompile Include="..\..\src\compare.c" />
    <ClCompile Include="..\..\src\connected_layer.c" />
    <ClCompile Include="..\..\src\convolutional_layer.c" />
    <ClCompile Include="..\..\src\conv_lstm_layer.c" />
    <ClCompile Include="..\..\src\cost_layer.c" />
    <ClCompile Include="..\..\src\cpu_gemm.c" />
    <ClCompile Include="..\..\src\crnn_layer.c" />
    <ClCompile Include="..\..\src\crop_layer.c" />
    <ClCompile Include="..\..\src\darknet.c" />
    <ClCompile Include="..\..\src\dark_cuda.c" />
    <ClCompile Include="..\..\src\data.c" />
    <ClCompile Include="..\..\src\deconvolutional_layer.c" />
    <ClCompile Include="..\..\src\demo.c" />
    <ClCompile Include="..\..\src\detection_layer.c" />
    <ClCompile Include="..\..\src\detector.c" />
    <ClCompile Include="..\..\src\dice.c" />
    <ClCompile Include="..\..\src\dropout_layer.c" />
    <ClCompile Include="..\..\src\gaussian_yolo_layer.c" />
    <ClCompile Include="..\..\src\gemm.c" />
    <ClCompile Include="..\..\src\getopt.c" />
    <ClCompile Include="..\..\src\gettimeofday.c" />
    <ClCompile Include="..\..\src\go.c" />
    <ClCompile Include="..\..\src\gru_layer.c" />
    <ClCompile Include="..\..\src\http_stream.cpp" />
    <ClCompile Include="..\..\src\im2col.c" />
    <ClCompile Include="..\..\src\image.c" />
    <ClCompile Include="..\..\src\image_opencv.cpp" />
    <ClCompile Include="..\..\src\layer.c" />
    <ClCompile Include="..\..\src\list.c" />
    <ClCompile Include="..\..\src\local_layer.c" />
    <ClCompile Include="..\..\src\lstm_layer.c" />
    <ClCompile Include="..\..\src\matrix.c" />
    <ClCompile Include="..\..\src\maxpool_layer.c" />
    <ClCompile Include="..\..\src\network.c" />
    <ClCompile Include="..\..\src\nightmare.c" />
    <ClCompile Include="..\..\src\normalization_layer.c" />
    <ClCompile Include="..\..\src\option_list.c" />
    <ClCompile Include="..\..\src\parser.c" />
    <ClCompile Include="..\..\src\region_layer.c" />
    <ClCompile Include="..\..\src\reorg_layer.c" />
    <ClCompile Include="..\..\src\reorg_old_layer.c" />
    <ClCompile Include="..\..\src\rnn.c" />
    <ClCompile Include="..\..\src\rnn_layer.c" />
    <ClCompile Include="..\..\src\rnn_vid.c" />
    <ClCompile Include="..\..\src\route_layer.c" />
    <ClCompile Include="..\..\src\sam_layer.c" />
    <ClCompile Include="..\..\src\scale_channels_layer.c" />
    <ClCompile Include="..\..\src\shortcut_layer.c" />
    <ClCompile Include="..\..\src\softmax_layer.c" />
    <ClCompile Include="..\..\src\super.c" />
    <ClCompile Include="..\..\src\swag.c" />
    <ClCompile Include="..\..\src\tag.c" />
    <ClCompile Include="..\..\src\tree.c" />
    <ClCompile Include="..\..\src\upsample_layer.c" />
    <ClCompile Include="..\..\src\utils.c" />
    <ClCompile Include="..\..\src\voxel.c" />
    <ClCompile Include="..\..\src\writing.c" />
    <ClCompile Include="..\..\src\yolo.c" />
    <ClCompile Include="..\..\src\yolo_layer.c" />
  </ItemGroup>
  <ItemGroup>
    <ClInclude Include="..\..\include\darknet.h" />
    <ClInclude Include="..\..\src\activations.h" />
    <ClInclude Include="..\..\src\activation_layer.h" />
    <ClInclude Include="..\..\src\avgpool_layer.h" />
    <ClInclude Include="..\..\src\batchnorm_layer.h" />
    <ClInclude Include="..\..\src\blas.h" />
    <ClInclude Include="..\..\src\box.h" />
    <ClInclude Include="..\..\src\classifier.h" />
    <ClInclude Include="..\..\src\col2im.h" />
    <ClInclude Include="..\..\src\connected_layer.h" />
    <ClInclude Include="..\..\src\convolutional_layer.h" />
    <ClInclude Include="..\..\src\conv_lstm_layer.h" />
    <ClInclude Include="..\..\src\cost_layer.h" />
    <ClInclude Include="..\..\src\crnn_layer.h" />
    <ClInclude Include="..\..\src\crop_layer.h" />
    <ClInclude Include="..\..\src\dark_cuda.h" />
    <ClInclude Include="..\..\src\data.h" />
    <ClInclude Include="..\..\src\deconvolutional_layer.h" />
    <ClInclude Include="..\..\src\demo.h" />
    <ClInclude Include="..\..\src\detection_layer.h" />
    <ClInclude Include="..\..\src\dropout_layer.h" />
    <ClInclude Include="..\..\src\gaussian_yolo_layer.h" />
    <ClInclude Include="..\..\src\gemm.h" />
    <ClInclude Include="..\..\src\getopt.h" />
    <ClInclude Include="..\..\src\gettimeofday.h" />
    <ClInclude Include="..\..\src\gru_layer.h" />
    <ClInclude Include="..\..\src\http_stream.h" />
    <ClInclude Include="..\..\src\im2col.h" />
    <ClInclude Include="..\..\src\image.h" />
    <ClInclude Include="..\..\src\image_opencv.h" />
    <ClInclude Include="..\..\src\layer.h" />
    <ClInclude Include="..\..\src\list.h" />
    <ClInclude Include="..\..\src\local_layer.h" />
    <ClInclude Include="..\..\src\lstm_layer.h" />
    <ClInclude Include="..\..\src\matrix.h" />
    <ClInclude Include="..\..\src\maxpool_layer.h" />
    <ClInclude Include="..\..\src\network.h" />
    <ClInclude Include="..\..\src\normalization_layer.h" />
    <ClInclude Include="..\..\src\option_list.h" />
    <ClInclude Include="..\..\src\parser.h" />
    <ClInclude Include="..\..\src\region_layer.h" />
    <ClInclude Include="..\..\src\reorg_layer.h" />
    <ClInclude Include="..\..\src\reorg_old_layer.h" />
    <ClInclude Include="..\..\src\rnn_layer.h" />
    <ClInclude Include="..\..\src\route_layer.h" />
    <ClInclude Include="..\..\src\sam_layer.h" />
    <ClInclude Include="..\..\src\scale_channels_layer.h" />
    <ClInclude Include="..\..\src\shortcut_layer.h" />
    <ClInclude Include="..\..\src\softmax_layer.h" />
    <ClInclude Include="..\..\src\stb_image.h" />
    <ClInclude Include="..\..\src\stb_image_write.h" />
    <ClInclude Include="..\..\src\tree.h" />
    <ClInclude Include="..\..\src\unistd.h" />
    <ClInclude Include="..\..\src\upsample_layer.h" />
    <ClInclude Include="..\..\src\utils.h" />
    <ClInclude Include="..\..\src\yolo_layer.h" />
  </ItemGroup>
  <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
  <ImportGroup Label="ExtensionTargets" />
</Project>

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
  • 15
  • 16
  • 17
  • 18
  • 19
  • 20
  • 21
  • 22
  • 23
  • 24
  • 25
  • 26
  • 27
  • 28
  • 29
  • 30
  • 31
  • 32
  • 33
  • 34
  • 35
  • 36
  • 37
  • 38
  • 39
  • 40
  • 41
  • 42
  • 43
  • 44
  • 45
  • 46
  • 47
  • 48
  • 49
  • 50
  • 51
  • 52
  • 53
  • 54
  • 55
  • 56
  • 57
  • 58
  • 59
  • 60
  • 61
  • 62
  • 63
  • 64
  • 65
  • 66
  • 67
  • 68
  • 69
  • 70
  • 71
  • 72
  • 73
  • 74
  • 75
  • 76
  • 77
  • 78
  • 79
  • 80
  • 81
  • 82
  • 83
  • 84
  • 85
  • 86
  • 87
  • 88
  • 89
  • 90
  • 91
  • 92
  • 93
  • 94
  • 95
  • 96
  • 97
  • 98
  • 99
  • 100
  • 101
  • 102
  • 103
  • 104
  • 105
  • 106
  • 107
  • 108
  • 109
  • 110
  • 111
  • 112
  • 113
  • 114
  • 115
  • 116
  • 117
  • 118
  • 119
  • 120
  • 121
  • 122
  • 123
  • 124
  • 125
  • 126
  • 127
  • 128
  • 129
  • 130
  • 131
  • 132
  • 133
  • 134
  • 135
  • 136
  • 137
  • 138
  • 139
  • 140
  • 141
  • 142
  • 143
  • 144
  • 145
  • 146
  • 147
  • 148
  • 149
  • 150
  • 151
  • 152
  • 153
  • 154
  • 155
  • 156
  • 157
  • 158
  • 159
  • 160
  • 161
  • 162
  • 163
  • 164
  • 165
  • 166
  • 167
  • 168
  • 169
  • 170
  • 171
  • 172
  • 173
  • 174
  • 175
  • 176
  • 177
  • 178
  • 179
  • 180
  • 181
  • 182
  • 183
  • 184
  • 185
  • 186
  • 187
  • 188
  • 189
  • 190
  • 191
  • 192
  • 193
  • 194
  • 195
  • 196
  • 197
  • 198
  • 199
  • 200
  • 201
  • 202
  • 203
  • 204
  • 205
  • 206
  • 207
  • 208
  • 209
  • 210
  • 211
  • 212
  • 213
  • 214
  • 215
  • 216
  • 217
  • 218
  • 219
  • 220
  • 221
  • 222
  • 223
  • 224
  • 225
  • 226
  • 227
  • 228
  • 229
  • 230
  • 231
  • 232
  • 233
  • 234
  • 235
  • 236
  • 237
  • 238
  • 239
  • 240
  • 241
  • 242
  • 243
  • 244
  • 245
  • 246
  • 247
  • 248
  • 249
  • 250
  • 251
  • 252
  • 253
  • 254
  • 255
  • 256
  • 257
  • 258
  • 259
  • 260
  • 261
  • 262
  • 263
  • 264
  • 265
  • 266
  • 267
  • 268
  • 269
  • 270
  • 271
  • 272
  • 273
  • 274
  • 275
  • 276
  • 277
  • 278
  • 279
  • 280
  • 281
  • 282
  • 283
  • 284
  • 285
  • 286
  • 287
  • 288
  • 289
  • 290
  • 291
  • 292
  • 293
  • 294
  • 295
  • 296
  • 297
  • 298
  • 299
  • 300
  • 301
  • 302
  • 303

封装及调用——方便工程化。

Windows系统下YOLO动态链接库的封装和调用(Windows10+VS2015+OpenCV3.4.0+CUDA9.0+cuDNN7.0)

https://blog.csdn.net/weixin_42783784/article/details/105741982

https://blog.csdn.net/Schling/article/details/114597758

其他参考

官方网址：https://github.com/hgpvision/darknet

2、YOLO3 darknet训练自己的数据

详细步骤！使用yolov3-tiny训练，测试、验证VOC数据集

声明：本文内容由网友自发贡献，不代表【wpsshop博客】立场，版权归原作者所有，本站不承担相应法律责任。如您发现有侵权的内容，请联系我们。转载请注明出处：https://www.wpsshop.cn/w/菜鸟追梦旅行/article/detail/380006
推荐阅读
相关标签

Copyright © 2003-2013 www.wpsshop.cn 版权所有，并保留所有权利。

  

闽ICP备14008679号