第五篇 学习OpenCV之视频处理
// 作者:GongchuangSu ( http://blog.csdn.net/gongchuangsu )
// 时间:2015.07.26
// 说明:从视频文件中读入数据,将读入数据转换为灰度图,对图像做Canny边缘检测。
// 将上面三个步骤实现显示在一个图像中,并在图像的三个不同部分写上合适的
// 标签,同时在这幅图像中任何地方点击鼠标,将当前图像的坐标显示于点击鼠标处。
// 注意:当R = G = B时为灰度色。
#include "cv.h"
#include "highgui.h"
#include <iostream>
using namespace std;
//#pragma comment(linker, "/subsystem:\"windows\" /entry:\"mainCRTStartup\"")
int col = 0;
int row = 0;
char *str = (char *)malloc(30*sizeof(char *));
// 定义 Canny 边缘检测函数
IplImage* doCanny(
IplImage* in,
int lowThresh,
int highThresh,
int aperture
){
if( in->nChannels != 1) // Canny只能处理灰度图片
return(0);
IplImage* out = cvCreateImage(
cvGetSize( in ),
in->depth,
in->nChannels
);
cvCanny( in, out, lowThresh, highThresh, aperture );
return( out );
};
// 定义鼠标回调函数
void my_mouse_callback( int event, int x, int y, int flags, void* param){
if( event == CV_EVENT_LBUTTONDOWN){
col = x;
row = y;
}
}
int main()
{
// 定义 Canny 边缘检测的阈值
int thresh = 30;
// 定义字体相关属性
CvFont font;
cvInitFont( &font, CV_FONT_HERSHEY_PLAIN, 1, 1, 0, 1, 8);
const char *AviName="MV.mp4";
const char *WindowsTitle = "Merge";
cvNamedWindow( WindowsTitle, 1);
CvCapture *capture=cvCreateFileCapture(AviName);
// 取得视频文件的大小并输出
CvSize size = cvSize(
(int)cvGetCaptureProperty( capture, CV_CAP_PROP_FRAME_WIDTH),
(int)cvGetCaptureProperty( capture, CV_CAP_PROP_FRAME_HEIGHT)
);
// 创建图像
IplImage *src = cvCreateImage( size, IPL_DEPTH_8U, 3);
IplImage *dst_gray = cvCreateImage( size, IPL_DEPTH_8U, 1);
IplImage *dst_canny = cvCreateImage( size, IPL_DEPTH_8U, 1);
IplImage *dst_merge = cvCreateImage( cvSize( size.width*3, size.height), IPL_DEPTH_8U, 3);
// 创建三个图像头,分别指向图像 dst_merge 的开始处、1/3处、2/3处
IplImage *src_header = cvCreateImageHeader( size, IPL_DEPTH_8U, 3);
src_header->origin = dst_merge->origin;
src_header->widthStep = dst_merge->widthStep;
src_header->imageData = dst_merge->imageData;
IplImage *dst_gray_header = cvCreateImageHeader( size, IPL_DEPTH_8U, 3);
dst_gray_header->origin = dst_merge->origin;
dst_gray_header->widthStep = dst_merge->widthStep;
dst_gray_header->imageData = dst_merge->imageData + dst_merge->widthStep/3;
IplImage *dst_canny_header = cvCreateImageHeader( size, IPL_DEPTH_8U, 3);
dst_canny_header->origin = dst_merge->origin;
dst_canny_header->widthStep = dst_merge->widthStep;
dst_canny_header->imageData = dst_merge->imageData + dst_merge->widthStep*2/3;
// 创建图像头 dst_gray_header 的单通道 r ,g ,b 图像
IplImage *dst_gray_header_r = cvCreateImage( cvGetSize(dst_gray_header), dst_gray_header->depth, 1);
IplImage *dst_gray_header_g = cvCreateImage( cvGetSize(dst_gray_header), dst_gray_header->depth, 1);
IplImage *dst_gray_header_b = cvCreateImage( cvGetSize(dst_gray_header), dst_gray_header->depth, 1);
// 创建图像头 dst_canny_header 的单通道 r ,g ,b 图像
IplImage *dst_canny_header_r = cvCreateImage( cvGetSize(dst_canny_header), dst_canny_header->depth, 1);
IplImage *dst_canny_header_g = cvCreateImage( cvGetSize(dst_canny_header), dst_canny_header->depth, 1);
IplImage *dst_canny_header_b = cvCreateImage( cvGetSize(dst_canny_header), dst_canny_header->depth, 1);
// 读取视频并转换
while( (src = cvQueryFrame(capture)) != NULL )
{
// 将原始图像拷贝给指向合并图像开始处的图像头
cvCopy( src, src_header);
// 将原始图像转换为灰度图像
cvConvertImage( src, dst_gray);
// 将灰度图像头分割成红、绿、蓝三个单通道图像,并将灰度图像拷贝给三个单通道图像,然后再将三个单通道
// 合并,以使灰度图像头显示的是灰度图像。
cvZero(dst_gray_header);
cvSplit( dst_gray_header, dst_gray_header_r, dst_gray_header_g, dst_gray_header_b, 0);
cvCopy( dst_gray, dst_gray_header_r);
cvCopy( dst_gray, dst_gray_header_g);
cvCopy( dst_gray, dst_gray_header_b);
cvMerge( dst_gray_header_r, dst_gray_header_g, dst_gray_header_b, 0, dst_gray_header);
// 将灰度图像进行 Canny 边缘检测处理
dst_canny = doCanny( dst_gray, thresh, 3*thresh, 3);
cvZero(dst_canny_header);
cvSplit( dst_canny_header, dst_canny_header_r, dst_canny_header_g, dst_canny_header_b, 0);
cvCopy( dst_canny, dst_canny_header_r);
cvCopy( dst_canny, dst_canny_header_g);
cvCopy( dst_canny, dst_canny_header_b);
cvMerge( dst_canny_header_r, dst_canny_header_g, dst_canny_header_b, 0, dst_canny_header);
// 在合并图像的三个不同处显示相应的文字
cvPutText( dst_merge, "Color", cvPoint(315,25), &font, cvScalar( 255, 255, 255));
cvPutText( dst_merge, "Gray", cvPoint(315*3,25), &font, cvScalar( 255, 255, 255));
cvPutText( dst_merge, "Canny", cvPoint(315*5,25), &font, cvScalar( 255, 255, 255));
cvShowImage( WindowsTitle, dst_merge);
// 设置鼠标回调函数
cvSetMouseCallback(
WindowsTitle,
my_mouse_callback,
(void *)dst_merge
);
// 输出坐标值
cout<<"( x, y) = "<<"( "<<col<<", "<<row<<")"<<endl;
memset( str, 0, 30);
// 将格式化的数据写入字符串缓冲区
sprintf( str, "( %d, %d)",col,row);
// 将字符串显示在相应的图像位置
cvPutText( dst_merge, str, cvPoint(col,row), &font, cvScalar( 0, 0, 255));
cvShowImage(WindowsTitle,dst_merge);
// 等待按键触发
char c = cvWaitKey(33);
if( c == 27)
break;
}
// 释放资源,销毁窗口
cvReleaseCapture(&capture);
cvReleaseImage(&dst_merge);
cvReleaseImage(&src);
cvReleaseImage(&dst_gray);
cvReleaseImage(&dst_canny);
cvReleaseImageHeader(&src_header);
cvReleaseImageHeader(&dst_gray_header);
cvReleaseImageHeader(&dst_canny_header);
cvReleaseImage(&dst_gray_header_r);
cvReleaseImage(&dst_gray_header_g);
cvReleaseImage(&dst_gray_header_b);
cvReleaseImage(&dst_canny_header_r);
cvReleaseImage(&dst_canny_header_g);
cvReleaseImage(&dst_canny_header_b);
cvDestroyAllWindows();
return(0);
}
运行结果:
版权声明:本文为博主原创文章,未经博主允许不得转载。