Haar训练中的对象标记

Question

我正在使用Haar训练进行有关OpenCV中对象检测的项目。 我必须在图像中标记对象存在的所有位置。 因此，我需要一个打开每个图像的代码，以便可以标记区域。 与这里的类似：

#include <opencv/cv.h>
#include <opencv/cvaux.h>
#include <opencv/highgui.h>

// for filelisting
#include <stdio.h>
#include <sys/io.h>
// for fileoutput
#include <string>
#include <fstream>
#include <sstream>
#include <dirent.h>
#include <sys/types.h>

using namespace std;

IplImage* image=0;
IplImage* image2=0;
//int start_roi=0;
int roi_x0=0;
int roi_y0=0;
int roi_x1=0;
int roi_y1=0;
int numOfRec=0;
int startDraw = 0;
char* window_name="<SPACE>add <B>save and load next <ESC>exit";

string IntToString(int num)
{
    ostringstream myStream; //creates an ostringstream object
    myStream << num << flush;
    /*
    * outputs the number into the string stream and then flushes
    * the buffer (makes sure the output is put into the stream)
    */
    return(myStream.str()); //returns the string form of the stringstream object
};

void on_mouse(int event,int x,int y,int flag, void *param)
{
    if(event==CV_EVENT_LBUTTONDOWN)
    {
        if(!startDraw)
        {
            roi_x0=x;
            roi_y0=y;
            startDraw = 1;
        } else {
            roi_x1=x;
            roi_y1=y;
            startDraw = 0;
        }
    }
    if(event==CV_EVENT_MOUSEMOVE && startDraw)
    {

        //redraw ROI selection
        image2=cvCloneImage(image);
        cvRectangle(image2,cvPoint(roi_x0,roi_y0),cvPoint(x,y),CV_RGB(255,0,255),1);
        cvShowImage(window_name,image2);
        cvReleaseImage(&image2);
    }

}

int main(int argc, char** argv)
{
    char iKey=0;
    string strPrefix;
    string strPostfix;
    string input_directory;
    string output_file;

    if(argc != 3) {
        fprintf(stderr, "%s output_info.txt raw/data/directory/\n", argv[0]);
        return -1;
    } 

    input_directory = argv[2];
    output_file = argv[1];

    /* Get a file listing of all files with in the input directory */
    DIR    *dir_p = opendir (input_directory.c_str());
    struct dirent *dir_entry_p;

    if(dir_p == NULL) {
        fprintf(stderr, "Failed to open directory %s\n", input_directory.c_str());
        return -1;
    }

    fprintf(stderr, "Object Marker: Input Directory: %s  Output File: %s\n", input_directory.c_str(), output_file.c_str());

    //    init highgui
    cvAddSearchPath(input_directory);
    cvNamedWindow(window_name,1);
    cvSetMouseCallback(window_name,on_mouse, NULL);

    fprintf(stderr, "Opening directory...");
    //    init output of rectangles to the info file
    ofstream output(output_file.c_str());
    fprintf(stderr, "done.\n");

    while((dir_entry_p = readdir(dir_p)) != NULL)
    {
        numOfRec=0;

        if(strcmp(dir_entry_p->d_name, ""))
        fprintf(stderr, "Examining file %s\n", dir_entry_p->d_name);

        /* TODO: Assign postfix/prefix info */
        strPostfix="";
        //strPrefix=input_directory;
        strPrefix=dir_entry_p->d_name;
        //strPrefix+=bmp_file.name;
        fprintf(stderr, "Loading image %s\n", strPrefix.c_str());

        if((image=cvLoadImage(strPrefix.c_str(),1)) != 0)
        {

            //    work on current image
            do

    {
                cvShowImage(window_name,image);

                // used cvWaitKey returns:
                //    <B>=66        save added rectangles and show next image
                //    <ESC>=27        exit program
                //    <Space>=32        add rectangle to current image
                //  any other key clears rectangle drawing only
                iKey=cvWaitKey(0);
                switch(iKey)
                {

                case 27:

                        cvReleaseImage(&image);
                        cvDestroyWindow(window_name);
                        return 0;
                case 32:

                        numOfRec++;
                printf("   %d. rect x=%d\ty=%d\tx2h=%d\ty2=%d\n",numOfRec,roi_x0,roi_y0,roi_x1,roi_y1);
                //printf("   %d. rect x=%d\ty=%d\twidth=%d\theight=%d\n",numOfRec,roi_x1,roi_y1,roi_x0-roi_x1,roi_y0-roi_y1);
                        // currently two draw directions possible:
                        //        from top left to bottom right or vice versa
                        if(roi_x0<roi_x1 && roi_y0<roi_y1)
                        {

                            printf("   %d. rect x=%d\ty=%d\twidth=%d\theight=%d\n",numOfRec,roi_x0,roi_y0,roi_x1-roi_x0,roi_y1-roi_y0);
                            // append rectangle coord to previous line content
                            strPostfix+=" "+IntToString(roi_x0)+" "+IntToString(roi_y0)+" "+IntToString(roi_x1-roi_x0)+" "+IntToString(roi_y1-roi_y0);

                        }
                        else
                                                    //(roi_x0>roi_x1 && roi_y0>roi_y1)
                        {
                            printf(" hello line no 154\n");
                            printf("   %d. rect x=%d\ty=%d\twidth=%d\theight=%d\n",numOfRec,roi_x1,roi_y1,roi_x0-roi_x1,roi_y0-roi_y1);
                            // append rectangle coord to previous line content
                            strPostfix+=" "+IntToString(roi_x1)+" "+IntToString(roi_y1)+" "+IntToString(roi_x0-roi_x1)+" "+IntToString      (roi_y0-roi_y1);
        }

                        break;
                }
            }
            while(iKey!=66);

            {
            // save to info file as later used for HaarTraining:
            //    <rel_path>\bmp_file.name numOfRec x0 y0 width0 height0 x1 y1 width1 height1...
            if(numOfRec>0 && iKey==66)
            {
                //append line
                /* TODO: Store output information. */
                output << strPrefix << " "<< numOfRec << strPostfix <<"\n";

            cvReleaseImage(&image);
            }

         else 
        {
            fprintf(stderr, "Failed to load image, %s\n", strPrefix.c_str());
        }
    }

    }}

    output.close();
    cvDestroyWindow(window_name);
    closedir(dir_p);

    return 0;
}

我使用Visual Studio 10运行了上面的代码。当我运行上面的代码时，它正在打开加载所有图像的命令窗口（一个新的命名窗口也正在打开但正在获取，但是我无法选择该区域），并且然后关闭。

当我在VS10中使用无调试启动选项运行它时，我可以看到以下内容：

Examining file img (6).jpeg
Loading image img (6).jpeg
Examining file img (6).JPG
Loading image img (6).JPG
Examining file img (7).jpeg
Loading image img (7).jpeg
Examining file img (7).jpg
Loading image img (7).jpg
Examining file img (8).jpeg
Loading image img (8).jpeg
Examining file img (8).jpg
Loading image img (8).jpg
Examining file img (9).jpeg
Loading image img (9).jpeg
Examining file img (9).jpg
Loading image img (9).jpg
Examining file img 1.jpeg
Loading image img 1.jpeg

对象标记失败，无法加载图像

我认为它正在跳过：

if((image=cvLoadImage(strPrefix.c_str(),1)) != 0){.....}

部分代码

Answer 1

当前很少有用于Haar training对象标记工具。 我在此链接中使用该工具。 它对我来说非常有效。 如果要构建自己的工具，请遵循本教程。 我认为以任何方式进行Haar training时，都要浪费时间来开发用于捕获坐标的工具，因为目前有用于这些坐标的工具，这并不是Haar training的主要目标。 因为Haar training需要更多时间来培训和创建级联XML文件。 因此，最好集中精力进行培训。