[Libav-user] only seeing 1/3 of image after sws_scale

[system159]

Hi all! I'm trying to use ffmpeg to decode single I frames (at user given
intervals) from a transport stream. The problem comes after calling
avcodec_decode_video2, as it returns the image in a crazy format. I want
basic RGB24 so that the rest of my QT app can use it via QImage. So I call
sws_scale to convert it to RGB24. When I later on post my image, is shows up
like this:
http://i54.tinypic.com/6rtx78.jpg

Can anybody tell me what I'm doing wrong? Is it sws_scale, or the way I'm
reading the data out of the avframe into the QImage?


My code is as follows(hard coded because I'm using a fixed data set to get
it working the first time).


avcodec_init(); 
    avcodec_register_all();
    pCodecCtx = NULL;
    pCodec = NULL;
    pFrame = NULL;
    pFrameRGB = NULL;
    pSwsCtx = NULL;    
    pFrame = avcodec_alloc_frame();
    pFrameRGB = avcodec_alloc_frame();

		int frameFinished;
        uint8_t *buffer;
        int numBytes;

        pCodecCtx = avcodec_alloc_context();
        
        // Determine required buffer size and allocate buffer
        numBytes = avpicture_get_size(PIX_FMT_RGB24, 720,
            480);

        pCodec = avcodec_find_decoder(CODEC_ID_MPEG2VIDEO);

        // Open codec
        if(avcodec_open(pCodecCtx, pCodec)<0)
        {

            return -1; // Could not open codec
        }

        AVPacket packet;

        //Fill AVPacket
        packet.data = codedImageData;
        packet.size = cidActualSize;
        //Decode video
        avcodec_decode_video2(pCodecCtx, pFrame, &frameFinished,
            &packet);

        if(frameFinished)
        {
            // Determine required buffer size and allocate buffer
            numBytes=avpicture_get_size(PIX_FMT_RGB24, 720,
                480);

            buffer=(uint8_t*)malloc(numBytes);

            // Assign appropriate parts of buffer to image planes in
pFrameRGB
            avpicture_fill((AVPicture *)pFrameRGB, buffer, PIX_FMT_RGB24,
                720, 480);
            struct  SwsContext *pSWSContext = sws_getContext(720, 480,
PIX_FMT_YUV420P, 
                720, 480, PIX_FMT_RGB24, SWS_BILINEAR, 
                NULL, NULL, NULL);
            sws_scale(pSWSContext, pFrame->data, pFrame->linesize, 0,
                480, pFrameRGB->data, pFrameRGB->linesize);
            
            unsigned char *src = (unsigned char *)pFrameRGB->data[0];
            QImage tmpImg(QSize(pFrame->width, pFrame->height),
QImage::Format_RGB888);
            for (int y = 0; y < pFrame->height; y++)
            {
                QRgb *scanLine = (QRgb *)gd->iFrame.scanLine(y);
                for (int x = 0; x < pFrame->width; x++)
                {
                    tmpImg.setPixel(QPoint(x, y), qRgb(src[3*x], src[3*x+1],
src[3*x+2]));
                }
                src += pFrameRGB->linesize[0];
            }
            
            sws_freeContext(pSwsCtx);
            ResetImageData();
            free(buffer);
        }

        pCodec->close(pCodecCtx);


Thanks!

--
View this message in context: http://libav-users.943685.n4.nabble.com/only-seeing-1-3-of-image-after-sws-scale-tp3655408p3655408.html
Sent from the libav-users mailing list archive at Nabble.com.