FFmpeg YUV420轉換為BGR24例子

/*
將保存有YUV420像素格式的幀，轉換成BGR24像素格式，并且按照幀指定的尺寸進行縮放
*/
void VideoDecodec::ConvertYUVFrameToBGRFrame(AVFrame* pYUVFrame, AVFrame* pBGRFrame)
{
?int nBGRFrameSize = av_image_get_buffer_size(AV_PIX_FMT_BGR24, pBGRFrame->width, pBGRFrame->height, 1);
?uint8_t* pszBGRBuffer = (uint8_t*)av_malloc(nBGRFrameSize);

?//將pszBGRBuffer掛載在pBGRFrame幀的圖片緩存指針，需要手動刪除
?av_image_fill_arrays(pBGRFrame->data, pBGRFrame->linesize, pszBGRBuffer, AV_PIX_FMT_BGR24, pBGRFrame->width, pBGRFrame->height, 1);

?struct SwsContext *pSwsCtx = sws_getContext(pYUVFrame->width, pYUVFrame->height, AV_PIX_FMT_YUVJ420P,
??pBGRFrame->width, pBGRFrame->height, AV_PIX_FMT_BGR24,
??SWS_POINT, NULL, NULL, NULL);

?//注意需要將0填充srcSliceY,否則調用失敗
?sws_scale(pSwsCtx, pYUVFrame->data,
??pYUVFrame->linesize, 0, pYUVFrame->height,
??pBGRFrame->data, pBGRFrame->linesize);

?//釋放環境

?sws_freeContext(pSwsCtx);
}

進階

sws_scale提供如下的算法對圖像進行伸縮變換

/* values for the flags, the stuff on the command line is different */
#define SWS_FAST_BILINEAR???? 1
#define SWS_BILINEAR????????? 2
#define SWS_BICUBIC?????????? 4
#define SWS_X???????????????? 8
#define SWS_POINT????????? 0x10
#define SWS_AREA?????????? 0x20
#define SWS_BICUBLIN?????? 0x40
#define SWS_GAUSS????????? 0x80
#define SWS_SINC????????? 0x100
#define SWS_LANCZOS?????? 0x200
#define SWS_SPLINE??????? 0x400

目前需要對圖像像素進行壓縮，方便算法加快識別，每種算法生成的像素圖片尺寸差不多