本文主要介绍在OpenCV中如何以逐帧方式和任意帧方式读取一个视频文件。

一、视频信息的读取

OpenCV中,VideoCapture类用于视频文件的读取,其方法get用于视频信息的获取。

比如获取,视频的宽高,帧数,原始视频的帧频。

VideoCapture cap;
cap.open("./test.mp4");

if (!cap.isOpened())//如果视频不能正常打开则返回
		return 0;

int frameRate = static_cast<int>(cap.get(CV_CAP_PROP_FPS));  //帧率 x frames/s
int totalFrames = static_cast<int>(cap.get(CV_CAP_PROP_FRAME_COUNT)); //总帧数
int width = static_cast<int>(cap.get(CV_CAP_PROP_FRAME_WIDTH));  //帧宽度
int height = static_cast<int>(cap.get(CV_CAP_PROP_FRAME_HEIGHT)); //帧高度

cout << "视频宽度=" << width << endl;
cout << "视频高度=" << height << endl;
cout << "视频总帧数=" << totalFrames << endl;
cout << "帧率=" << frameRate << endl;

二、视频的逐帧去读

OpenCV中,从一个视频文件中按顺序逐帧获取图像是非常简单的,这里提供三种方法:

1、流操作 >> 方法

2、read()方法

3、grap()和retrieve()方法

   其实,read()方法就是将grab和retrieve整合到了一个调用函数中而已。

VideoCapture cap;
cap.open("./test.mp4");

if (!cap.isOpened())//如果视频不能正常打开则返回
		return 0;

Mat frame;

double t = 0, fps = 0;

while (1)
{
    t = (double)cv::getTickCount();

    //方法1:
    //cap >> frame;

    //方法2:
    //cap.read(frame);

    //方法3:
    cap.grab();
    cap.retrieve(frame);


	if (frame.empty())//如果某帧为空则退出循环
	    break;
	
	t = ((double)cv::getTickCount() - t) / cv::getTickFrequency();
	fps = 1.0 / t;

	cout << " 帧频是: " << fps << endl;

}

 

三、视频的任意帧获取

视频的任意位置获取,需要首先设置任意帧的位置,

VideoCaputure 通过其方法set设置要获取帧的位置,具体有以下几种方法:

set(CAP_PROP_POS_MSEC, 毫秒为单位的时间位置)

set(CAP_PROP_POS_FRAMES , 任意帧的索引位置)

set(CAP_PROP_POS_AVI_RATIO  , 任意帧的位置在整个视频中的位置比例, 数值为0-1)

 

比如:下面的例子是以6帧的间隔读取视频,不过亲测,读取速度较慢!

int timeStame = 0;

while (1)
	{
		
		timeStame += 6;


		t = (double)cv::getTickCount();
		cap.set(CAP_PROP_POS_FRAMES, timeStame);
		cap.read(frame);

		t = ((double)cv::getTickCount() - t) / cv::getTickFrequency();
		fps = 1.0 / t;

		cout << " 帧频是: " << fps << endl;


		//cap >> frame;//等价于cap.read(frame);

		if (frame.empty())//如果某帧为空则退出循环
			break;

	}

 

 

 

 

 

 

 

 

GitHub 加速计划 / opencv31 / opencv
166
15
下载
OpenCV: 开源计算机视觉库
最近提交(Master分支:4 个月前 )
c623a5af Android camera refactoring #26646 This patch set does not contain any functional changes. It just cleans up the code structure to improve readability and to prepare for future changes. * videoio(Android): Use 'unique_ptr' instead of 'shared_ptr' Using shared pointers for unshared data is considered an antipattern. * videoio(Android): Make callback functions private static members Don't leak internal functions into global namespace. Some member variables are now private as well. * videoio(Android): Move resolution matching into separate function Also make internally used member functions private. * videoio(Android): Move ranges query into separate function Also remove some unneccessary initialisations from initCapture(). * videoio(Android): Wrap extremly long source code lines * videoio(Android): Rename members of 'RangeValue' ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [x] I agree to contribute to the project under Apache 2 License. - [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [ ] The PR is proposed to the proper branch - [ ] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake 4 小时前
7728dd33 Fix potential READ memory access #26782 This fixes https://oss-fuzz.com/testcase-detail/4923671881252864 and https://oss-fuzz.com/testcase-detail/5048650127966208 ### Pull Request Readiness Checklist See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request - [x] I agree to contribute to the project under Apache 2 License. - [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV - [x] The PR is proposed to the proper branch - [x] There is a reference to the original bug report and related work - [ ] There is accuracy test, performance test and test data in opencv_extra repository, if applicable Patch to opencv_extra has the same branch name. - [ ] The feature is well documented and sample code can be built with the project CMake 6 小时前
Logo

旨在为数千万中国开发者提供一个无缝且高效的云端环境,以支持学习、使用和贡献开源项目。

更多推荐