OpenCV2.4 例程五 (人脸检测)

以下程序来自OpenCV自带例程:

原例是检测视频文件中的人脸，笔者改为检测视频中的人脸，还有OpenCV2.4的摄像头的驱动我该成了VFW，否则不能正常读取摄像头：

#include "stdafx.h"
#include "opencv/highgui.h"
#include "opencv2/objdetect/objdetect.hpp"
#include "opencv2/video/tracking.hpp"
#include "opencv2/video/video.hpp"
#include "opencv2/highgui/highgui.hpp"
#include "opencv2/imgproc/imgproc.hpp"

#include <iostream>
#include <stdio.h>

using namespace std;
using namespace cv;

void help()
{
cout << "\nThis program demonstrates the multi cascade recognizer. It is a generalization of facedetect sample.\n\n"
"Usage: ./multicascadeclassifier \n"
"   --cascade1=<cascade_path> this is the primary trained classifier such as frontal face\n"
"   [--cascade2=[this an optional secondary classifier such as profile face or eyes]]\n"
"   input video or image\n\n"
"Example: ./multicascadeclassifier --cascade1=\"../../data/haarcascades/haarcascade_frontalface_alt.xml\" --cascade2=\"../../data/haarcascades/haarcascade_eye.xml\"\n\n"
"Using OpenCV version " << CV_VERSION << endl << endl;
}

void DetectAndDraw(Mat& img, CascadeClassifier& cascade);

String cascadeName = "H:/OpenCV2.4/opencv/data/haarcascades/haarcascade_frontalface_alt.xml";
String nestedCascadeName = "H:/OpenCV2.4/opencv/data/haarcascades/haarcascade_eye_tree_eyeglasses.xml";

int main( int argc, const char** argv )
{
CvCapture* capture = 0;
//double t=capture->getProperty(0);
//capture = cvCreateCameraCapture(0);//cvCaptureFromCAM( 0 );//
Mat frame, image;

if (argc == 0)
{
help();
return 0;
}

const String cascadeOpt = "--cascade1=";
size_t cascadeOptLen = cascadeOpt.length();
string inputName;
for( int i = 1; i < argc; i++ )
{
cout << "Processing argument #" << i << ": " <<  argv[i] << endl;
if( cascadeOpt.compare( 0, cascadeOptLen, argv[i], cascadeOptLen ) == 0 )
{
cascadeName.assign( argv[i] + cascadeOptLen );
cout << "  from which we have cascadeName= " << cascadeName << endl;
}
else if( argv[i][0] == '-' )
{
cerr << "WARNING: Unknown option " << argv[i] << endl;
}
else
inputName.assign( argv[i] );
}

CascadeClassifier cascade;
if( !cascade.load( cascadeName ) )
{
cerr << "ERROR: Could not load cascade classifier \"" << cascadeName << "\"" << endl;
help();

return -1;
}

if( inputName.size() )
{
image = imread( inputName, 1 );
if( image.empty() )
{
capture = cvCaptureFromAVI( inputName.c_str() );
if( !capture )
cout << "Capture from AVI don't work" << endl;
}
}
else
{
capture=cvCaptureFromCAM(-1);
/*cout << "Please provide input file." << endl;
return -1;*/
}

cvNamedWindow( "result", 1 );

if( capture )
{
for(;;)
{
IplImage* iplImg = cvQueryFrame( capture );
frame = iplImg;
if( frame.empty() )
break;

DetectAndDraw( frame, cascade );

if( waitKey( 10 ) >= 0 )
goto _cleanup_;
}

waitKey(0);
_cleanup_:
cvReleaseCapture( &capture );
}
else if( !image.empty() )
{
DetectAndDraw( image, cascade );
waitKey(0);
}
else
{
cout << "Please provide correct input file." << endl;
}

cvDestroyWindow("result");

return 0;
}

void DetectAndDraw( Mat& img, CascadeClassifier& cascade)
{
int i = 0;
double t = 0;
vector<Rect> faces;
const static Scalar colors[] =  { CV_RGB(0,0,255),
CV_RGB(0,128,255),
CV_RGB(0,255,255),
CV_RGB(0,255,0),
CV_RGB(255,128,0),
CV_RGB(255,255,0),
CV_RGB(255,0,0),
CV_RGB(255,0,255)} ;
Mat gray;
Mat frame( cvRound(img.rows), cvRound(img.cols), CV_8UC1 );

cvtColor( img, gray, CV_BGR2GRAY );
resize( gray, frame, frame.size(), 0, 0, INTER_LINEAR );
equalizeHist( frame, frame );

t = (double)cvGetTickCount();
cascade.detectMultiScale( frame, faces,
1.1, 2, 0
//|CV_HAAR_FIND_BIGGEST_OBJECT
//|CV_HAAR_DO_ROUGH_SEARCH
|CV_HAAR_SCALE_IMAGE
,
Size(30, 30) );
t = (double)cvGetTickCount() - t;
printf( "detection time = %g ms\n", t/((double)cvGetTickFrequency()*1000.) );

for( vector<Rect>::const_iterator r = faces.begin(); r != faces.end(); r++, i++ )
{
Point center;
Scalar color = colors[i%8];
int radius;
center.x = cvRound(r->x + r->width*0.5);
center.y = cvRound(r->y + r->height*0.5);
radius = (int)(cvRound(r->width + r->height)*0.25);
circle( img, center, radius, color, 3, 8, 0 );
}

cv::imshow( "result", img );
}

缺点是检测效率还达不到实时，700ms左右每帧： 效果图：