使用javacv中的ffmpeg实现录屏，结果连运行都失败了，现在终于解决了

前言：

今天突发奇想，想自己写一个录屏的软件，上次写了一个专门录音的Demo，但是要把声音和视频放到一起合成一个mp4文件，着实有一点艰难，所以就打算使用ffmpeg来写一个，而这篇博客中会顺便谈一谈我碰到的各种坑。
ffmpeg是一个c++程序，要想在java中使用ffmpeg，无非就是两种方式：直接在java程序中调用ffmpeg.exe，还有就是通过jni的方式。而在这里我就是使用jni的方式，但是我在这里直接使用javacv这个框架来实现就可以，用这个的好处就是你什么都不要干，直接导入几个重要的jar包就可以。

步骤：

首先呢，下载javacv就可以：github下载
你也可以直接在我这里下载
然后下载好了，就要开始导包了，导入javacpp，javacv，这2个一定要导，另外要能使用ffmpeg的API实现录屏就要再导入
ffmpeg-windows-x86_64.jar
ffmpeg-platform.jar
ffmpeg.jar
videoinput.jar

导完包之后代码测试一下，这里发一个别人写的代码，可以实现录屏，但不能录音，代码里面需要修改一下存放文件的路径：

package com;

import java.awt.AWTException;
import java.awt.Graphics2D;
import java.awt.Rectangle;
import java.awt.RenderingHints;
import java.awt.Robot;
import java.awt.Toolkit;
import java.awt.image.BufferedImage;
import java.nio.ByteBuffer;
import java.nio.ByteOrder;
import java.nio.ShortBuffer;
import java.util.Scanner;
import java.util.concurrent.ScheduledThreadPoolExecutor;
import java.util.concurrent.TimeUnit;

import javax.sound.sampled.AudioFormat;
import javax.sound.sampled.AudioSystem;
import javax.sound.sampled.DataLine;
import javax.sound.sampled.LineUnavailableException;
import javax.sound.sampled.TargetDataLine;

import org.bytedeco.ffmpeg.global.avcodec;
import org.bytedeco.ffmpeg.global.avutil;

import org.bytedeco.javacv.FFmpegFrameRecorder;
import org.bytedeco.javacv.Frame;
import org.bytedeco.javacv.FrameRecorder.Exception;
import org.bytedeco.javacv.Java2DFrameConverter;

	/**
 * 使用javacv进行录屏
 * 
 */
public class VideoRecord {
	
	//线程池 screenTimer
	private ScheduledThreadPoolExecutor screenTimer;
	//获取屏幕尺寸
	private final Rectangle rectangle = new Rectangle(Constant.WIDTH, Constant.HEIGHT); // 截屏的大小
	//视频类 FFmpegFrameRecorder
	private FFmpegFrameRecorder recorder;
	private Robot robot;
	//线程池 exec
	private ScheduledThreadPoolExecutor exec;
	private TargetDataLine line;
	private AudioFormat audioFormat;
	private DataLine.Info dataLineInfo;
	private boolean isHaveDevice = true;
	private long startTime = 0;
	private long videoTS = 0;
	private long pauseTime = 0;
	private double frameRate = 5;

	public VideoRecord(String fileName, boolean isHaveDevice) {
		// TODO Auto-generated constructor stub
		recorder = new FFmpegFrameRecorder(fileName + ".mp4", Constant.WIDTH, Constant.HEIGHT);
		// recorder.setVideoCodec(avcodec.AV_CODEC_ID_H265); // 28
		// recorder.setVideoCodec(avcodec.AV_CODEC_ID_FLV1); // 28
		recorder.setVideoCodec(avcodec.AV_CODEC_ID_MPEG4); // 13
		recorder.setFormat("mp4");
		// recorder.setFormat("mov,mp4,m4a,3gp,3g2,mj2,h264,ogg,MPEG4");
		recorder.setSampleRate(44100);
		recorder.setFrameRate(frameRate);
		recorder.setVideoQuality(0);
		recorder.setVideoOption("crf", "23");
		// 2000 kb/s, 720P视频的合理比特率范围
		recorder.setVideoBitrate(1000000);
		/**
		 * 权衡quality(视频质量)和encode speed(编码速度) values(值)： ultrafast(终极快),superfast(超级快),
		 * veryfast(非常快), faster(很快), fast(快), medium(中等), slow(慢), slower(很慢),
		 * veryslow(非常慢)
		 * ultrafast(终极快)提供最少的压缩（低编码器CPU）和最大的视频流大小；而veryslow(非常慢)提供最佳的压缩（高编码器CPU）的同时降低视频流的大小
		 * 参考：https://trac.ffmpeg.org/wiki/Encode/H.264 官方原文参考：-preset ultrafast as the
		 * name implies provides for the fastest possible encoding. If some tradeoff
		 * between quality and encode speed, go for the speed. This might be needed if
		 * you are going to be transcoding multiple streams on one machine.
		 */
		recorder.setVideoOption("preset", "slow");
		recorder.setPixelFormat(avutil.AV_PIX_FMT_YUV420P); // yuv420p
		recorder.setAudioChannels(2);
		recorder.setAudioOption("crf", "0");
		// Highest quality
		recorder.setAudioQuality(0);
		recorder.setAudioCodec(avcodec.AV_CODEC_ID_AAC);
		try {
			robot = new Robot();
		} catch (AWTException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
		try {
			recorder.start();
		} catch (Exception e) {
			// TODO Auto-generated catch block
			System.out.print("*******************************");
		}
		this.isHaveDevice = isHaveDevice;
	}

	/**
	 * 开始录制
	 */
	public void start() {

		if (startTime == 0) {
			startTime = System.currentTimeMillis();
		}
		if (pauseTime == 0) {
			pauseTime = System.currentTimeMillis();
		}
		// 如果有录音设备则启动录音线程
		if (isHaveDevice) {
			new Thread(new Runnable() {

				@Override
				public void run() {
					// TODO Auto-generated method stub
					caputre();
				}
			}).start();

		}

		// 录屏
		screenTimer = new ScheduledThreadPoolExecutor(1);
		screenTimer.scheduleAtFixedRate(new Runnable() {
			@Override
			public void run() {

				// 将screenshot对象写入图像文件
				// try {
				// ImageIO.write(screenCapture, "JPEG", f);
				// videoGraphics.drawImage(screenCapture, 0, 0, null);
				// IplImage image = cvLoadImage(name); // 非常吃内存！！
				// // 创建一个 timestamp用来写入帧中
				// videoTS = 1000
				// * (System.currentTimeMillis() - startTime - (System.currentTimeMillis() -
				// pauseTime));
				// // 检查偏移量
				// if (videoTS > recorder.getTimestamp()) {
				// recorder.setTimestamp(videoTS);
				// }
				BufferedImage screenCapture = robot.createScreenCapture(rectangle); // 截屏
				
				BufferedImage videoImg = new BufferedImage(Constant.WIDTH, Constant.HEIGHT,
						BufferedImage.TYPE_3BYTE_BGR); // 声明一个BufferedImage用重绘截图
				
				Graphics2D videoGraphics = videoImg.createGraphics();// 创建videoImg的Graphics2D
				
				videoGraphics.setRenderingHint(RenderingHints.KEY_DITHERING, RenderingHints.VALUE_DITHER_DISABLE);
				videoGraphics.setRenderingHint(RenderingHints.KEY_COLOR_RENDERING,
						RenderingHints.VALUE_COLOR_RENDER_SPEED);
				videoGraphics.setRenderingHint(RenderingHints.KEY_RENDERING, RenderingHints.VALUE_RENDER_SPEED);
				videoGraphics.drawImage(screenCapture, 0, 0, null); // 重绘截图
				
				Java2DFrameConverter java2dConverter = new Java2DFrameConverter();
				
				Frame frame = java2dConverter.convert(videoImg);
				try {
					videoTS = 1000L
							* (System.currentTimeMillis() - startTime - (System.currentTimeMillis() - pauseTime));
					// 检查偏移量
					if (videoTS > recorder.getTimestamp()) {
						recorder.setTimestamp(videoTS);
					}
					recorder.record(frame); // 录制视频
				} catch (Exception e) {
					// TODO Auto-generated catch block
					e.printStackTrace();
				}
				// 释放资源
				videoGraphics.dispose();
				videoGraphics = null;
				videoImg.flush();
				videoImg = null;
				java2dConverter = null;
				screenCapture.flush();
				screenCapture = null;

			}
		}, (int) (1000 / frameRate), (int) (1000 / frameRate), TimeUnit.MILLISECONDS);

	}

	/**
	 * 抓取声音
	 */
	public void caputre() {
		audioFormat = new AudioFormat(44100.0F, 16, 2, true, false);
		dataLineInfo = new DataLine.Info(TargetDataLine.class, audioFormat);
		try {
			line = (TargetDataLine) AudioSystem.getLine(dataLineInfo);
		} catch (LineUnavailableException e1) {
			// TODO Auto-generated catch block
			System.out.println("#################");
		}
		try {
			line.open(audioFormat);
		} catch (LineUnavailableException e1) {
			// TODO Auto-generated catch block
			e1.printStackTrace();
		}
		line.start();

		final int sampleRate = (int) audioFormat.getSampleRate();
		final int numChannels = audioFormat.getChannels();

		int audioBufferSize = sampleRate * numChannels;
		final byte[] audioBytes = new byte[audioBufferSize];

		exec = new ScheduledThreadPoolExecutor(1);
		exec.scheduleAtFixedRate(new Runnable() {
			@Override
			public void run() {
				try {
					int nBytesRead = line.read(audioBytes, 0, line.available());
					int nSamplesRead = nBytesRead / 2;
					short[] samples = new short[nSamplesRead];

					// Let's wrap our short[] into a ShortBuffer and
					// pass it to recordSamples
					ByteBuffer.wrap(audioBytes).order(ByteOrder.LITTLE_ENDIAN).asShortBuffer().get(samples);
					ShortBuffer sBuff = ShortBuffer.wrap(samples, 0, nSamplesRead);

					// recorder is instance of
					// org.bytedeco.javacv.FFmpegFrameRecorder
					recorder.recordSamples(sampleRate, numChannels, sBuff);
					// System.gc();
				} catch (org.bytedeco.javacv.FrameRecorder.Exception e) {
					e.printStackTrace();
				}
			}
		}, (int) (1000 / frameRate), (int) (1000 / frameRate), TimeUnit.MILLISECONDS);
	}

	/**
	 * 停止
	 */
	public void stop() {
		if (null != screenTimer) {
			screenTimer.shutdownNow();
		}
		try {
			recorder.stop();
			recorder.release();
			recorder.close();
			screenTimer = null;
			// screenCapture = null;
			if (isHaveDevice) {
				if (null != exec) {
					exec.shutdownNow();
				}
				if (null != line) {
					line.stop();
					line.close();
				}
				dataLineInfo = null;
				audioFormat = null;
			}
		} catch (Exception e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}

	}

	/**
	 * 暂停
	 * 
	 * @throws Exception
	 */
	public void pause() throws Exception {
		screenTimer.shutdownNow();
		screenTimer = null;
		if (isHaveDevice) {
			exec.shutdownNow();
			exec = null;
			line.stop();
			line.close();
			dataLineInfo = null;
			audioFormat = null;
			line = null;
		}
		pauseTime = System.currentTimeMillis();
	}

	public static void main(String[] args) throws Exception, AWTException {
		VideoRecord videoRecord = new VideoRecord("C:\\Users\\Administrator\\Desktop\\视频", false);
		videoRecord.start();
		while (true) {
			System.out.println("你要停止吗？请输入(stop)，程序会停止。");
			Scanner sc = new Scanner(System.in);
			if (sc.next().equalsIgnoreCase("stop")) {
				videoRecord.stop();
				System.out.println("停止");
			}
			if (sc.next().equalsIgnoreCase("pause")) {
				videoRecord.pause();
				System.out.println("暂停");
			}
			if (sc.next().equalsIgnoreCase("start")) {
				videoRecord.start();
				System.out.println("开始");
			}
		}
	}
}

class Constant{
	public final static int WIDTH=Toolkit.getDefaultToolkit().getScreenSize().width;
	public final static int HEIGHT=Toolkit.getDefaultToolkit().getScreenSize().height;

}

如果你能成功stop并保存一个视频文件，那就说明可以使用ffmpeg了。

我碰到的各种坑：

1）import的时候，一些函数什么的找不到定义。像这两个，在我copy的代码中是org.bytedeco.javacpp.avcodec和org.bytedeco.javacpp.avutil，但包里根本找不到，后面找了找发现是在ffmpeg.jar包下，所以要改成
import org.bytedeco.ffmpeg.global.avcodec;
import org.bytedeco.ffmpeg.global.avutil;
2）少导入了一些jar包，因为我看其他人的博客，只要导入javacpp，javacv就可以，这是最让人崩溃的，搞得我一直找不到到底是哪的问题，后来一股脑全给导进来又可以了，然后一步步去排除，才找到了哪些jar包。因为这个我找了几天的博客，差点就放弃了。

下一篇博客，是对上面代码的完善：javacv利用ffmpeg实现录屏和录音，输出为mp4文件