ffmpeg超詳細綜合教程——攝像頭直播

本文的示例將實現：讀取PC攝像頭視頻數據並以RTMP協議發送為直播流。示例包含了
1、ffmpeg的libavdevice的使用
2、視頻解碼、編碼、推流的基本流程
具有較強的綜合性。
要使用libavdevice的相關函數，首先需要注冊相關組件

avdevice_register_all();

接下來我們要列出電腦中可用的dshow設備

AVFormatContext *pFmtCtx = avformat_alloc_context();
    AVDeviceInfoList *device_info = NULL;
    AVDictionary* options = NULL;
    av_dict_set(&options, "list_devices", "true", 0);
    AVInputFormat *iformat = av_find_input_format("dshow");
    printf("Device Info=============\n");
    avformat_open_input(&pFmtCtx, "video=dummy", iformat, &options);
    printf("========================\n");

可以看到這里打開設備的步驟基本與打開文件的步驟相同，上面的代碼中設置了AVDictionary，這樣與在命令行中輸入下列命令有相同的效果

ffmpeg -list_devices true -f dshow -i dummy

以上語句得到的結果如下

這里我的電腦上只有一個虛擬攝像頭軟件虛擬出來的幾個dshow設備，沒有音頻設備，所以有如上的結果。
需要說明的是，avdevice有一個avdevice_list_devices函數可以枚舉系統的采集設備，包括設備名和設備描述，非常適合用於讓用戶選擇要使用的設備，但是不支持dshow設備，所以這里沒有使用它。
下一步就可以像打開普通文件一樣將上面的具體設備名作為輸入打開，並進行相應的初始化設置，如下

av_register_all();
    //Register Device
    avdevice_register_all();
    avformat_network_init();

    //Show Dshow Device
    show_dshow_device();

    printf("\nChoose capture device: ");
    if (gets(capture_name) == 0)
    {
        printf("Error in gets()\n");
        return -1;
    }
    sprintf(device_name, "video=%s", capture_name);

    ifmt=av_find_input_format("dshow");

    //Set own video device's name
    if (avformat_open_input(&ifmt_ctx, device_name, ifmt, NULL) != 0){
        printf("Couldn't open input stream.（無法打開輸入流）\n");
        return -1;
    }
    //input initialize
    if (avformat_find_stream_info(ifmt_ctx, NULL)<0)
    {
        printf("Couldn't find stream information.（無法獲取流信息）\n");
        return -1;
    }
    videoindex = -1;
    for (i = 0; i<ifmt_ctx->nb_streams; i++)
        if (ifmt_ctx->streams[i]->codec->codec_type == AVMEDIA_TYPE_VIDEO)
        {
            videoindex = i;
            break;
        }
    if (videoindex == -1)
    {
        printf("Couldn't find a video stream.（沒有找到視頻流）\n");
        return -1;
    }
    if (avcodec_open2(ifmt_ctx->streams[videoindex]->codec, avcodec_find_decoder(ifmt_ctx->streams[videoindex]->codec->codec_id), NULL)<0)
    {
        printf("Could not open codec.（無法打開解碼器）\n");
        return -1;
    }

在選擇了輸入設備並進行相關初始化之后，需要對輸出做相應的初始化。ffmpeg將網絡協議和文件同等看待，同時因為使用RTMP協議進行傳輸，這里我們指定輸出為flv格式，編碼器使用H.264

//output initialize
    avformat_alloc_output_context2(&ofmt_ctx, NULL, "flv", out_path);
    //output encoder initialize
    pCodec = avcodec_find_encoder(AV_CODEC_ID_H264);
    if (!pCodec){
        printf("Can not find encoder! (沒有找到合適的編碼器！)\n");
        return -1;
    }
    pCodecCtx=avcodec_alloc_context3(pCodec);
    pCodecCtx->pix_fmt = PIX_FMT_YUV420P;
    pCodecCtx->width = ifmt_ctx->streams[videoindex]->codec->width;
    pCodecCtx->height = ifmt_ctx->streams[videoindex]->codec->height;
    pCodecCtx->time_base.num = 1;
    pCodecCtx->time_base.den = 25;
    pCodecCtx->bit_rate = 400000;
    pCodecCtx->gop_size = 250;
    /* Some formats,for example,flv, want stream headers to be separate. */
    if (ofmt_ctx->oformat->flags & AVFMT_GLOBALHEADER)
        pCodecCtx->flags |= CODEC_FLAG_GLOBAL_HEADER;

    //H264 codec param
    //pCodecCtx->me_range = 16;
    //pCodecCtx->max_qdiff = 4;
    //pCodecCtx->qcompress = 0.6;
    pCodecCtx->qmin = 10;
    pCodecCtx->qmax = 51;
    //Optional Param
    pCodecCtx->max_b_frames = 3;
    // Set H264 preset and tune
    AVDictionary *param = 0;
    av_dict_set(¶m, "preset", "fast", 0);
    av_dict_set(¶m, "tune", "zerolatency", 0);

    if (avcodec_open2(pCodecCtx, pCodec,¶m) < 0){
        printf("Failed to open encoder! (編碼器打開失敗！)\n");
        return -1;
    }

    //Add a new stream to output,should be called by the user before avformat_write_header() for muxing
    video_st = avformat_new_stream(ofmt_ctx, pCodec);
    if (video_st == NULL){
        return -1;
    }
    video_st->time_base.num = 1;
    video_st->time_base.den = 25;
    video_st->codec = pCodecCtx;

    //Open output URL,set before avformat_write_header() for muxing
    if (avio_open(&ofmt_ctx->pb,out_path, AVIO_FLAG_READ_WRITE) < 0){
    printf("Failed to open output file! (輸出文件打開失敗！)\n");
    return -1;
    }

    //Show some Information
    av_dump_format(ofmt_ctx, 0, out_path, 1);

    //Write File Header
    avformat_write_header(ofmt_ctx,NULL);

完成輸入和輸出的初始化之后，就可以正式開始解碼和編碼並推流的流程了，這里要注意，攝像頭數據往往是RGB格式的，需要將其轉換為YUV420P格式，所以要先做如下的准備工作

//prepare before decode and encode
    dec_pkt = (AVPacket *)av_malloc(sizeof(AVPacket));
    //enc_pkt = (AVPacket *)av_malloc(sizeof(AVPacket));
    //camera data has a pix fmt of RGB,convert it to YUV420
    img_convert_ctx = sws_getContext(ifmt_ctx->streams[videoindex]->codec->width, ifmt_ctx->streams[videoindex]->codec->height,
        ifmt_ctx->streams[videoindex]->codec->pix_fmt, pCodecCtx->width, pCodecCtx->height, PIX_FMT_YUV420P, SWS_BICUBIC, NULL, NULL, NULL);
    pFrameYUV = avcodec_alloc_frame();
    uint8_t *out_buffer = (uint8_t *)av_malloc(avpicture_get_size(PIX_FMT_YUV420P, pCodecCtx->width, pCodecCtx->height));
    avpicture_fill((AVPicture *)pFrameYUV, out_buffer, PIX_FMT_YUV420P, pCodecCtx->width, pCodecCtx->height);

下面就可以正式開始解碼、編碼和推流了

//start decode and encode
    int64_t start_time=av_gettime();
    while (av_read_frame(ifmt_ctx, dec_pkt) >= 0){
        if (exit_thread)
            break;
        av_log(NULL, AV_LOG_DEBUG, "Going to reencode the frame\n");
        pframe = av_frame_alloc();
        if (!pframe) {
            ret = AVERROR(ENOMEM);
            return -1;
        }
        //av_packet_rescale_ts(dec_pkt, ifmt_ctx->streams[dec_pkt->stream_index]->time_base,
        //  ifmt_ctx->streams[dec_pkt->stream_index]->codec->time_base);
        ret = avcodec_decode_video2(ifmt_ctx->streams[dec_pkt->stream_index]->codec, pframe,
            &dec_got_frame, dec_pkt);
        if (ret < 0) {
            av_frame_free(&pframe);
            av_log(NULL, AV_LOG_ERROR, "Decoding failed\n");
            break;
        }
        if (dec_got_frame){
            sws_scale(img_convert_ctx, (const uint8_t* const*)pframe->data, pframe->linesize, 0, pCodecCtx->height, pFrameYUV->data, pFrameYUV->linesize);

            enc_pkt.data = NULL;
            enc_pkt.size = 0;
            av_init_packet(&enc_pkt);
            ret = avcodec_encode_video2(pCodecCtx, &enc_pkt, pFrameYUV, &enc_got_frame);
            av_frame_free(&pframe);
            if (enc_got_frame == 1){
                //printf("Succeed to encode frame: %5d\tsize:%5d\n", framecnt, enc_pkt.size);
                framecnt++;
                enc_pkt.stream_index = video_st->index;

                //Write PTS
                AVRational time_base = ofmt_ctx->streams[videoindex]->time_base;//{ 1, 1000 };
                AVRational r_framerate1 = ifmt_ctx->streams[videoindex]->r_frame_rate;// { 50, 2 };
                AVRational time_base_q = { 1, AV_TIME_BASE };
                //Duration between 2 frames (us)
                int64_t calc_duration = (double)(AV_TIME_BASE)*(1 / av_q2d(r_framerate1));  //內部時間戳
                //Parameters
                //enc_pkt.pts = (double)(framecnt*calc_duration)*(double)(av_q2d(time_base_q)) / (double)(av_q2d(time_base));
                enc_pkt.pts = av_rescale_q(framecnt*calc_duration, time_base_q, time_base);
                enc_pkt.dts = enc_pkt.pts;
                enc_pkt.duration = av_rescale_q(calc_duration, time_base_q, time_base); //(double)(calc_duration)*(double)(av_q2d(time_base_q)) / (double)(av_q2d(time_base));
                enc_pkt.pos = -1;

                //Delay
                int64_t pts_time = av_rescale_q(enc_pkt.dts, time_base, time_base_q);
                int64_t now_time = av_gettime() - start_time;
                if (pts_time > now_time)
                    av_usleep(pts_time - now_time);

                ret = av_interleaved_write_frame(ofmt_ctx, &enc_pkt);
                av_free_packet(&enc_pkt);
            }
        }
        else {
            av_frame_free(&pframe);
        }
        av_free_packet(dec_pkt);
    }

解碼部分比較簡單，編碼部分需要自己計算PTS、DTS，比較復雜。這里通過幀率計算PTS和DTS
首先通過幀率計算每兩幀之間的時間間隔，但是要換算