ffmpeg解复用视频文件

from: http://blog.chinaunix.net/uid-24922718-id-3692670.html

解复用视频文件：就是提取出音频文件和视频文件的ES流数据。而对于aac和h264来说，提取出来的数据直接保存并不能成功的播放，原因就在ffmpeg解出的ES数据比较纯净，没有ADTS头，而很多解码器都是需要ADTS头信息来进行解码的，所以就要对aac数据进行头部adts结构的添加。
     对于h264文件来说，h264有两种封装，
一种是annexb模式，传统模式，有startcode，SPS和PPS是在ES中
一种是mp4模式，一般mp4 mkv会有，没有startcode，SPS和PPS以及其它信息被封装在container中，每一个frame前面是这个frame的长度
很多解码器只支持annexb这种模式，因此需要将mp4做转换：
在ffmpeg中用h264_mp4toannexb_filter可以做转换
实现：
注册filter
avcbsfc = av_bitstream_filter_init("h264_mp4toannexb");
转换bitstream
av_bitstream_filter_filter(AVBitStreamFilterContext *bsfc,
AVCodecContext *avctx, const char *args,
uint8_t **poutbuf, int *poutbuf_size,
const uint8_t *buf, int buf_size, int keyframe)

核心代码：

点击(此处)折叠或打开

 #include <libavformat/avformat.h>
 #include <libavcodec/avcodec.h>
 #include <stdio.h>
 static int audioindex = -1;
 static int videoindex = -1;
 static int isaaccodec = -1;
 int main(int argc ,char **argv)
 {
     av_register_all();
     FILE *f = NULL;
     FILE *g = NULL;
     f = fopen("audio","wb");
     g = fopen("video","wb");
     if(!f || !g)
     {
         printf("open write file errorn");
         return 0;
     }
     AVFormatContext *fmtctx = NULL;

     AVPacket audiopack;
     if(avformat_open_input(&fmtctx,argv[1],NULL,NULL) < 0)
     {
         printf("open fmtctx errorn");
         return 0;
     }

     if(avformat_find_stream_info(fmtctx,NULL) < 0)
     {
         printf("find stream info n");
         return 0;
     }
     int streamnum = fmtctx->nb_streams;
     printf("stream num is %dn",streamnum);
     int i=0;
     for(;i<streamnum;i++)
     {
         if(fmtctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_AUDIO && audioindex == -1)
         {
             audioindex = i;
         }
         else if(fmtctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO && videoindex == -1)
         {
             videoindex = i;
         }
     }
     printf("audioindex is %dn",audioindex);

     printf("videoindex is %dn",videoindex);

     AVCodecContext *codecctx = fmtctx->streams[videoindex]->codec;
     AVCodec *decode = avcodec_find_decoder(codecctx->codec_id);

     AVCodecContext *audioCodecCtx = fmtctx->streams[audioindex]->codec;
     AVCodec *audiodecode = avcodec_find_decoder(audioCodecCtx->codec_id);
     if(audiodecode->id == AV_CODEC_ID_AAC)
     {
         isaaccodec = 1;
     }

     if(avcodec_open2(codecctx,decode,NULL) < 0)
     {
         return -1;
     }
     if(avcodec_open2(audioCodecCtx,audiodecode,NULL) < 0)
     {
         return -1;
     }
     printf("extradata size is %dn",audioCodecCtx->extradata_size);

     AVBitStreamFilterContext* bsfc = av_bitstream_filter_init("h264_mp4toannexb");
     AVBitStreamFilterContext* aacbsfc = av_bitstream_filter_init("aac_adtstoasc");
     if(!bsfc || !aacbsfc)
     {
         return 0;
     }
     AVFrame picture;
     while(!(av_read_frame(fmtctx,&audiopack)))
     {
         if(audiopack.stream_index == 1)
         {
             if(isaaccodec == 1)
             {
                 char bits[7] = {0};
                 int sample_index = 0 , channel = 0;
                 char temp = 0;
                 int length = 7 + audiopack.size;
                 sample_index = (audioCodecCtx->extradata[0] & 0x07) << 1;
                 temp = (audioCodecCtx->extradata[1]&0x80);
                 switch(audioCodecCtx->sample_rate)
                 {
                     case 44100:
                         {
                             sample_index = 0x7;
                         }break;
                     default:
                         {
                             sample_index = sample_index + (temp>>7);
                         }break;
                 }
                 channel = ((audioCodecCtx->extradata[1] - temp) & 0xff) >> 3;
                 bits[0] = 0xff;
                 bits[1] = 0xf1;
                 bits[2] = 0x40 | (sample_index<<2) | (channel>>2);
                 bits[3] = ((channel&0x3)<<6) | (length >>11);
                 bits[4] = (length>>3) & 0xff;
                 bits[5] = ((length<<5) & 0xff) | 0x1f;
                 bits[6] = 0xfc;

                 fwrite(bits,1,7,f);
             }
             fwrite(audiopack.data,1,audiopack.size,f);
             printf("audio pts is %fn",audiopack.pts*av_q2d(fmtctx->streams[audioindex]->time_base));
         }
         else if(audiopack.stream_index == videoindex){
             AVPacket pkt = audiopack;

             int a = av_bitstream_filter_filter(bsfc, codecctx, NULL, &pkt.data, &pkt.size, audiopack.data, audiopack.size, audiopack.flags& AV_PKT_FLAG_KEY);
             if(a > 0)
             {
                 audiopack = pkt;
             }
             fwrite(audiopack.data,1,audiopack.size,g);
             int gotfinished = 0;
             if(avcodec_decode_video2(codecctx,&picture,&gotfinished,&audiopack) < 0)
             {
                 printf("decode video errorn");
             }
             if(picture.key_frame)
             {
                 //printf("key framen");
             }
             else{
                 //printf("frame num is %dn",picture.pict_type);
             }

             //printf("video pts is %f, %dn",picture.pkt_dts * av_q2d(fmtctx->streams[videoindex]->codec->time_base),picture.pts );
         }
         av_free_packet(&audiopack);
     }
     fclose(f);
 }

最后就可以用vlc 或ffplay来查看解析出来的音频aac文件和264文件，本套代码是通用解析器，任何视频文件都可以解析，只是对于264和aac单独做了处理。
贴上最后的效果图：