ffmpeg save rtsp stream


#include <stdio.h>


#ifdef __cplusplus

extern "C" {

#endif

#include <libavcodec/avcodec.h>

#include <libavformat/avformat.h>

//#include <libswscale/swscale.h>


#ifdef _MSC_VER

int strcasecmp(const char *s1, const char *s2)

{

    while ((*s1 != ‘\0‘)

        && (tolower(*(unsigned char *) s1) ==

        tolower(*(unsigned char *) s2)))

    {

        s1++;

        s2++;

    }


return tolower(*(unsigned char *) s1) - tolower(*(unsigned char *) s2);

}

int strncasecmp(const char *s1, const char *s2, unsigned int n)

{

    if (n == 0)

        return 0;


while ((n-- != 0)

        && (tolower(*(unsigned char *) s1) ==

        tolower(*(unsigned char *) s2))) {

            if (n == 0 || *s1 == ‘\0‘ || *s2 == ‘\0‘)

                return 0;

            s1++;

            s2++;

    }


return tolower(*(unsigned char *) s1) - tolower(*(unsigned char *) s2);

}

#endif //_MSC_VER


#ifdef __cplusplus

}

#endif


/***********************************************************

‘** stream_component_open

*    Description:

*             //open the stream component for video/audio

*

*

*    Params:ic-pointer which contains the url and codec details

:stream_index-index denoting video or audio stream

***********************************************************/

int stream_component_open(AVFormatContext *ic, int stream_index)

{

    AVCodecContext *enc;

    AVCodec *codec;


if (stream_index < 0 || stream_index >= (int)ic->nb_streams)

        return -1;

    enc = ic->streams[stream_index]->codec;


/* prepare audio output */

    if (enc->codec_type == CODEC_TYPE_AUDIO)

    {

        if (enc->channels > 0)

            enc->request_channels = FFMIN(2, enc->channels);

        else

            enc->request_channels = 2;


/*Hardcoding the codec id to PCM_MULAW if the audio

        codec id returned by the lib is CODEC_ID_NONE */


if(enc->codec_id == CODEC_ID_NONE)

        {

            enc->codec_id = CODEC_ID_PCM_MULAW;

            enc->channels = 1;

            enc->sample_rate = 16000;

            enc->bit_rate = 128;

        }

    }


codec = avcodec_find_decoder(enc->codec_id);

    enc->idct_algo            = FF_IDCT_AUTO;

    enc->flags2   |= CODEC_FLAG2_FAST;

    enc->skip_frame            = AVDISCARD_DEFAULT;

    enc->skip_idct            = AVDISCARD_DEFAULT;

    enc->skip_loop_filter    = AVDISCARD_DEFAULT;

    enc->error_concealment    = 3;


if (!codec || avcodec_open(enc, codec) < 0)

        return -1;

    avcodec_thread_init(enc, 1);

    enc->thread_count= 1;

    ic->streams[stream_index]->discard = AVDISCARD_DEFAULT;


return 0;

}

//声明函数

int enable_local_record(AVFormatContext *ic/*已经打开的视频文件上下文*/,

                        int videostream,int audiostream,

                        AVFormatContext **out_oc,const char * ofile,const char * ofileformat);


int exit_onerr(const char * err_desc/*错误描述*/,int err_code/*错误码*/)

{

    printf("%s\n",err_desc);

    system("pause");//暂停，查看错误描述

    return err_code;

}

int main(int argc, char* argv[])

{

    //初始化ffmpeg链表结构

    avcodec_register_all();                        // Register all formats and codecs

    av_register_all();


AVPacket  packet;


//打开文件

    AVFormatContext * ic = NULL;

    const char * rtsp_url = "rtsp://192.168.0.168:8557/PSIA/Streaming/channels/2?videoCodecType=H.264";

    if(av_open_input_file(&ic, rtsp_url, NULL, 0, NULL)!=0)

    {

        return exit_onerr("can‘t open file.",-1);

    }

    if(!ic)

    {

        return exit_onerr("unknow error.",-2);

    }

    ic ->max_analyze_duration = 1000;


//get streams information

    if(av_find_stream_info(ic)<0)

    {

        av_close_input_file(ic);//退出前，记得释放资源

        ic = NULL;

        return exit_onerr("con‘t init streams information.",-3);

    }


//find stream

    int videoStream=-1;    // Didn‘t find a video stream

    int audioStream=-1;    // Didn‘t find a audio stream

    // Find the first video stream

    for(int i=0; i<ic ->nb_streams; i++)

    {

        if( ic ->streams[i]->codec->codec_type==CODEC_TYPE_VIDEO)

        {

            videoStream=i;

            break;

        }

    }

    // Find the first audio stream

    for(int i=0; i<ic ->nb_streams; i++)

    {

        if( ic ->streams[i]->codec->codec_type==CODEC_TYPE_AUDIO)

        {

            audioStream=i;

            break;

        }

    }

    //判断视频文件中是否包含音视频流，如果没有，退出

    if(audioStream<0 && videoStream<0)

    {

        av_close_input_file(ic);//退出前，记得释放资源

        ic = NULL;

        return exit_onerr("con‘t find a audio stream or video stream",-4);

    }


//open the stream component for video

    int videoComponent = -1;

    if(videoStream >= 0)

    {

        videoComponent = stream_component_open(ic, videoStream);

        if(videoComponent<0)

        {

            av_close_input_file(ic);//退出前，记得释放资源

            return exit_onerr("not supported video stream",-5);//要求重新编译ffmpeg以支持该种编码格式

        }

    }

    //open the stream component for audio

    int audioComponent = -1;

    if(audioStream >= 0)

    {

        audioComponent = stream_component_open(ic, audioStream);

        if(audioComponent<0)

        {

            av_close_input_file(ic);//退出前，记得释放资源

            return exit_onerr("not supported audio stream",-6);//要求重新编译ffmpeg以支持该种编码格式

        }

    }


//////////////////////////////////////////////////////

    int ret = 0;

    //初始化并打开录像

    AVFormatContext * oc = NULL;

    const char * out_file_name = "D:\\test.avi";


//获取一帧完整的图像用于初始化ffmpeg内部的一些结构体数据(这里使用什么数据尚未清楚，请自行查看)，否则会出现写文件头失败

    int got_picture = 0;

    AVFrame *frame = avcodec_alloc_frame();

    while( 1 )

    {

        if(av_read_frame( ic, &packet)<0)

        {

            av_free_packet(&packet);

            av_close_input_file(ic);//退出前，记得释放资源

            return exit_onerr("read frame error.",-7);//读取视频帧失败

        }

        if(packet.stream_index == videoStream)

        {

            avcodec_decode_video(ic->streams[videoStream]->codec, frame, &got_picture, packet.data, packet.size);

        }

        av_free_packet(&packet);

        if(got_picture)

            break;

    }

    av_free(frame);


ret = enable_local_record(ic,videoStream,audioStream,&oc,out_file_name,"avi");


if(ret <0 || !oc)

    {

        //退出前，记得释放资源，现在又多了个oc参数需要释放

        if(oc)

        {

            ///cleanup the output contents format

            for(unsigned int i=0;i< oc->nb_streams;i++)

            {

                av_metadata_free(&oc ->streams[i]->metadata);

                av_free(oc ->streams[i]->codec);

                av_free(oc ->streams[i]);

            }

            for(unsigned int i=0;i<oc ->nb_programs;i++)

            {

                av_metadata_free(&oc ->programs[i]->metadata);

            }

            for(unsigned int i=0;i<oc ->nb_chapters;i++)

            {

                av_metadata_free(&oc ->chapters[i]->metadata);

            }

            av_metadata_free(&oc ->metadata);

            av_free(oc);

        }

        av_close_input_file(ic);

        return exit_onerr("can‘t init out file.",-8);

    }


//开始录像

    int video_dts = 0,audio_dts = 0;//时间戳

    int total_frame = 300;//写300帧文件

    while(total_frame--)

    {

        if( av_read_frame( ic, &packet) <0 )        //read the packet

        {

            //读取数据出错

            av_free_packet(&packet);

            break;

        }

        if(packet.data && (packet.stream_index == videoStream || packet.stream_index == audioStream) )

        {

            //计算时间视频戳，顺序+1，这里可以多研究几种编码的音视频文件，求其时间戳生成格式，h264编码顺序加1即可

            if(packet.stream_index == videoStream)

            {

                packet.dts = video_dts++;

                packet.pts = video_dts;

            }

            else if(packet.stream_index == audioStream)//计算音频时间戳

            {

                packet.dts = audio_dts++;

                packet.pts = audio_dts * (1000 * packet.size /ic ->streams[packet.stream_index]->codec ->sample_rate);

            }

            packet.flags |= PKT_FLAG_KEY;

            if(av_interleaved_write_frame(oc,&packet)<0)

            {

                printf("st:%d\twrite frame failed.\n",packet.stream_index);

            }

        }

        av_free_packet(&packet);

    }


//关闭录像文件和输入文件


//写文件尾

    av_write_trailer(oc);


/* close the output file if need.*/

    if (!(oc ->oformat->flags & AVFMT_NOFILE))

    {

        url_fclose(oc->pb);

    }


//释放资源

    /*cleanup the output contents format*/

    for(unsigned int i=0;i< oc->nb_streams;i++)

    {

        av_metadata_free(&oc ->streams[i]->metadata);

        av_free(oc ->streams[i]->codec);

        av_free(oc ->streams[i]);

    }

    for(unsigned int i=0;i<oc ->nb_programs;i++)

    {

        av_metadata_free(&oc ->programs[i]->metadata);

    }

    for(unsigned int i=0;i<oc ->nb_chapters;i++)

    {

        av_metadata_free(&oc ->chapters[i]->metadata);

    }

    av_metadata_free(&oc ->metadata);

    av_free(oc);


av_close_input_file(ic);


return 0;

}


//初始化录像文件，代码较长，可写成函数

/*

*return <0  failed, 0 success

*/

int enable_local_record(AVFormatContext *ic/*已经打开的视频文件上下文*/,int videostream,int audiostream,AVFormatContext **out_oc,const char * ofile,const char * ofileformat)

{

    AVFormatContext *oc        = NULL;//

    AVOutputFormat  *fmt    = NULL;


if( ofileformat )

    {

        fmt = av_guess_format(ofileformat, NULL, NULL);

    }

    if(!fmt)

    {

        printf("out file format \"%s\" invalidate.\n",ofileformat);

        return -1;

    }


/*******************************************init output contents begin**********************************************************/

    /* allocate the output media context */

    oc = avformat_alloc_context();

    if (!oc)

    {

        printf("out of memory.\n");

        return -2;//内存分配失败

    }


*out_oc = oc;


oc ->oformat = fmt;

    sprintf_s( oc ->filename, sizeof( oc ->filename), "%s", ofile);

    av_metadata_conv(oc, fmt->metadata_conv, NULL);

    if( videostream >=0 )

    {

        //AVCodecContext* pCodecCtx= ->streams[videostream]->codec;;

        //add video stream

        AVStream * st = av_new_stream(oc, videostream);

        if(!st)

        {

            printf("can not add a video stream.\n");

            return -3;

        }

        st->codec->codec_id            =  ic ->streams[videostream] ->codec->codec_id;

        st->codec->codec_type        =  CODEC_TYPE_VIDEO;

        st->codec->bit_rate            =  ic ->streams[videostream] ->codec->bit_rate;

        st->codec->width            =  ic ->streams[videostream] ->codec->width;

        st->codec->height            =  ic ->streams[videostream] ->codec->height;

        st->codec->gop_size            =  ic ->streams[videostream] ->codec->gop_size;

        st->codec->pix_fmt            =  ic ->streams[videostream] ->codec->pix_fmt;

        st->codec->frame_size        =  ic ->streams[videostream] ->codec->frame_size;

        st->codec->has_b_frames        =  ic ->streams[videostream] ->codec->has_b_frames;

        st->codec->extradata        =  ic ->streams[videostream] ->codec->extradata;

        st->codec->extradata_size    =  ic ->streams[videostream] ->codec->extradata_size;

        st->codec->codec_tag        =  ic ->streams[videostream] ->codec->codec_tag;

        st->codec->bits_per_raw_sample        =  ic ->streams[videostream] ->codec->bits_per_raw_sample;

        st->codec->chroma_sample_location    =  ic ->streams[videostream] ->codec->chroma_sample_location;

        st->time_base.den            =  ic ->streams[videostream] ->time_base.den;

        st->time_base.num            =  ic ->streams[videostream] ->time_base.num;

        st->cur_dts                    =  ic ->streams[videostream] ->cur_dts;

        st->stream_copy                =  1;

        st->pts.den                    =  ic ->streams[videostream] ->time_base.den;

        st->pts.num                    =  ic ->streams[videostream] ->time_base.num;

        if( oc ->oformat->flags & AVFMT_GLOBALHEADER)

            st ->codec->flags |= CODEC_FLAG_GLOBAL_HEADER;

        if(av_q2d(ic ->streams[videostream] ->codec->time_base)*ic ->streams[videostream] ->codec->ticks_per_frame > av_q2d(ic ->streams[videostream]->time_base) &&

            av_q2d(ic ->streams[videostream]->time_base) < 1.0/1000)

        {

            st->codec->time_base = ic ->streams[videostream] ->codec->time_base;

            st->codec->time_base.num *= ic ->streams[videostream] ->codec->ticks_per_frame;

        }

        else

        {

            st->codec->time_base = ic ->streams[videostream] ->time_base;

        }

        st->disposition                =  ic ->streams[videostream] ->disposition;

    }

    if(audiostream >= 0 )

    {

        AVStream * st = av_new_stream( oc, audiostream);

        if(!st)

        {

            printf("can not add a audio stream.\n");

            return -4;

        }

        st->codec->codec_id            =  ic ->streams[audiostream] ->codec->codec_id;

        st->codec->codec_type        =  CODEC_TYPE_AUDIO;

        st->codec->bit_rate            =  ic ->streams[audiostream] ->codec->bit_rate;

        st->codec->gop_size            =  ic ->streams[audiostream] ->codec->gop_size;

        st->codec->pix_fmt            =  ic ->streams[audiostream] ->codec->pix_fmt;

        st->codec->bit_rate            =  ic ->streams[audiostream] ->codec->bit_rate;

        st->codec->channel_layout    =  ic ->streams[audiostream] ->codec->channel_layout;

        st->codec->frame_size        =  ic ->streams[audiostream] ->codec->frame_size;

        st->codec->sample_rate        =  ic ->streams[audiostream] ->codec->sample_rate;

        st->codec->channels            =  ic ->streams[audiostream] ->codec->channels;

        st->codec->block_align        =  ic ->streams[audiostream] ->codec->block_align;

        st->time_base.den            =  ic ->streams[audiostream] ->time_base.den;

        st->time_base.num            =  ic ->streams[audiostream] ->time_base.num;

        st->stream_copy                =  1;

        st->pts.den                    =  ic ->streams[audiostream] ->time_base.den;

        st->pts.num                    =  ic ->streams[audiostream] ->time_base.num;

        if( oc->oformat->flags & AVFMT_GLOBALHEADER)

            st ->codec->flags |= CODEC_FLAG_GLOBAL_HEADER;

        if(av_q2d(ic ->streams[audiostream] ->codec->time_base)*ic ->streams[audiostream] ->codec->ticks_per_frame > av_q2d(ic ->streams[audiostream]->time_base) &&

            av_q2d(ic ->streams[audiostream]->time_base) < 1.0/1000)

        {

            st->codec->time_base = ic ->streams[audiostream] ->codec->time_base;

            st->codec->time_base.num *= ic ->streams[audiostream] ->codec->ticks_per_frame;

        }

        else

        {

            st->codec->time_base = ic ->streams[audiostream] ->time_base;

        }

    }

    /* set the output parameters (must be done even if no parameters). */

    //AVFormatParameters params, *ap = ?ms;

    //memset(ap, 0, sizeof(*ap));

    if (av_set_parameters(oc, NULL/*ap*/) < 0)

    {

        printf("invalid output format parameters.\n");

        return -5;

    }

    oc ->flags |= AVFMT_FLAG_NONBLOCK;

    /*******************************************init output contents end**********************************************************/


/* open the output file, if needed */

    if (!(oc ->oformat ->flags & AVFMT_NOFILE))

    {

        try

        {

            if (url_fopen(&oc->pb, ofile, URL_WRONLY) < 0)

            {

                printf("Could not open file.\n");

                return -6;

            }

        }

        catch(...)

        {

            printf("Could not open file.\n");

            return -6;

        }

    }


/* write the stream header, if any */

    if( 0 != av_write_header(oc))

    {

        printf("write the stream header failed.\n");

        return -7;

    }


return 0;

}

ffmpeg save rtsp stream,布布扣,bubuko.com

时间： 2024-10-14 10:46:14

ffmpeg save rtsp stream的相关文章

(原)关于获取ffmpeg解析rtsp流sdp中带有sps,pps的情况

 转载请注明出处:http://www.cnblogs.com/lihaiping/p/6612511.html 今天同事准备在android下使用ffmpeg来获取rtsp流,问我如何获取获取sps和pps的情况?因为尽管使用ffmpeg做项目还算熟悉,但用ffmpeg作rtsp client我还真没用过,虽然知道ffmpeg支持rtsp,但一般rtsp client我还是选择自己实现.使用ffmpeg做rtsp client的在这种情况下,如果rtsp的pps,sps在sdp中带有的话,我们

ffmpeg接收rtsp流问题

项目使用mingw环境g++5.3,C++调用ffmpeg接收rtsp流,再通过C#显示.结构上是C#调用C++的so文件,读取得到的视频帧(RGB888格式),通过图片控件显示. 一开始是使用opencv打开视频源,本地文件和rtsp的源使用一样的接口,方便使用.但是通过opencv打开rtsp源的时候,发现720p2Mbit的视频能够正常打开,但1080p的视频打开后出马赛克,卡顿比较多. 而同样的视频源,使用VLC就能流畅的打开,不过延时很高,大约1s. 后来又发现,使用i3低电压版CPU

基于FFmpeg的RTSP directshow filter

常规来讲,directshow filter包含三个部分,分别为source filter.transform filter.renderfilter 因为项目需要,要开发一个基于FFmpeg的filter(soucer filter+transform filter合并).即source filter -> render filter 此filter需要能解码和显示rtsp的h264视频流.大致方案是基于ffmpeg,利用ffmpeg接收rtsp视频流,获取每一个avframe并将其转换为RG

在Centos 7.x中使用ffmpeg搭建RTSP流处理环境

Centos7.x 搭建RTSP流处理环境 服务器环境 服务器 centos 7 目前版本CentOS Linux release 7.6.1810下载地址 安装nginx & nginx-rtmp-module 官网下载nginx 目前版本nginx-1.17.0下载地址 下载依赖pcre 目前版本pcre-8.38下载地址 使用pcre2 nginx会编译失败 下载依赖zlib 目前版本zlib-1.2.11下载地址 下载openssl 目前版本openssl-1.0.2s下载地址 下载ng

ffmpeg解码RTSP/TCP视频流H.264(QT界面显示视频画面)

我用的ffmpeg版本为 ffmpeg-2.1.8.tar.bz2 版本低了恐怕有些头文件和API找不到. 在Linux下解压后编译,Linux下编译很简单,我这里生成的动态库: ./configure –enable-shared make 就能找到各个so动态库文件. 移动位置后,记得手动链接 一下: ln -s libavcodec.so.55 libavcodec.so ln -s libavdevice.so.55 libavdevice.so ln -s libavfilter.so

ffmpeg之rtsp分析流程

1,首先从ffplay.c main() is = stream_open(input_filename, file_iformat); 2,在straem_open里面 is->read_tid = SDL_CreateThread(read_thread, is); 3,read_thread err = avformat_open_input(&ic, is->filename, is->iformat, &format_opts); 4,avformat_open

ffmpeg+rtsp+dss

1. push stream to dss ffmpeg -f mpegts -re -i film.v -c:v libx264 -s 352x288 -aspect 4:3 -b:v 300k -c:a libfaac  -ac 1  -f rtsp 'rtsp://127.0.0.1/live1.sdp' 2. play stream from dss ffmpeg -i 'rtsp://test:[email protected]:554/live1.sdp'  -an -c:v raw

PHP+ffmpeg RTSP视频流转换成RTMP视频流

目前360浏览器可以支持VLC插件之外,其他浏览器和高版本的 Chrome 浏览器不支持 NPAPI 插件,也就是说高版本的 Chrome 浏览器还是不能播放(46 以上的版本都不行). PHP+NGINX+FFMPEG 转换流媒体协议代码 1 <?php 2 //PHP不接受执行命令后的信息,执行命令后立刻执行其他PHP程序 3 //$str = "ffmpeg -i RTSP地址 -vcodec copy -acodec copy -f flv 转换后的RTMP服务器地址 >/d

ffmpeg推送rtsp码流到rtmp服务器

HIKVISION DS-2CD23RZ admin 12345 rtsp://admin:[email protected]:554/H.264/ch1/main/av_stream ffmpeg -i "rtsp://admin:[email protected]:554/H.264/ch1/main/av_stream" -f flv -r 25 -s 1920x1080 -an "rtmp://182.61.45.149:1935/live/movie" f