javacv实现视频截帧功能

引入依赖

        <dependency>
            <groupId>org.bytedeco</groupId>
            <artifactId>javacv</artifactId>
            <version>0.8</version>
        </dependency>

使用

调用方法：InterceptVideoPhotoUtil.fetchFrame(源视频文件路径, 截取帧的图片存放路径);

public class InterceptVideoPhotoUtil {
    
    

    /**
     * 获取指定视频的帧并保存为图片至指定目录
     *
     * @param videofile 源视频文件路径
     * @param framefile 截取帧的图片存放路径
     * @throws Exception▏
     */
    public static void fetchFrame(String videofile, String framefile)
            throws Exception {
    
    
        long start = System.currentTimeMillis();
        File targetFile = new File(framefile);
        FFmpegFrameGrabber ff = new FFmpegFrameGrabber(videofile);
        ff.start();
        int lenght = ff.getLengthInFrames();
        int i = 0;
        Frame f = null;
        while (i < lenght) {
    
    
            // 过滤前5帧，避免出现全黑的图片，依自己情况而定
            f = ff.grabFrame();
            if ((i > 5) && (f.image != null)) {
    
    
                break;
            }
            i++;
        }
        IplImage img = f.image;
        int owidth = img.width();
        int oheight = img.height();
        // 对截取的帧进行等比例缩放
        int width = 800;
        int height = (int) (((double) width / owidth) * oheight);
        BufferedImage bi = new BufferedImage(width, height, BufferedImage.TYPE_3BYTE_BGR);
        bi.getGraphics().drawImage(f.image.getBufferedImage().getScaledInstance(width, height, Image.SCALE_SMOOTH),
                0, 0, null);
        ImageIO.write(bi, "jpg", targetFile);
        //ff.flush();
        ff.stop();
        //System.out.println(System.currentTimeMillis() - start);
    }
}