[FFmpeg-cvslog] avfilter/vf_dnn_processing.c: add planar yuv format support
Guo, Yejun
git at videolan.org
Thu Mar 12 12:32:44 EET 2020
ffmpeg | branch: master | Guo, Yejun <yejun.guo at intel.com> | Fri Feb 21 14:20:48 2020 +0800| [bd50453894182af095c7d7578596e6ff6c58f852] | committer: Guo, Yejun
avfilter/vf_dnn_processing.c: add planar yuv format support
Only the Y channel is handled by dnn, the UV channels are copied
without changes.
The command to use srcnn.pb (see vf_sr) looks like:
./ffmpeg -i 480p.jpg -vf format=yuv420p,scale=w=iw*2:h=ih*2,dnn_processing=dnn_backend=tensorflow:model=srcnn.pb:input=x:output=y -y srcnn.jpg
Signed-off-by: Guo, Yejun <yejun.guo at intel.com>
Reviewed-by: Pedro Arthur <bygrandao at gmail.com>
> http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=bd50453894182af095c7d7578596e6ff6c58f852
---
doc/filters.texi | 9 ++++++
libavfilter/vf_dnn_processing.c | 72 +++++++++++++++++++++++++++++++++++++++++
2 files changed, 81 insertions(+)
diff --git a/doc/filters.texi b/doc/filters.texi
index f988ea4c80..554eaff163 100644
--- a/doc/filters.texi
+++ b/doc/filters.texi
@@ -9196,6 +9196,8 @@ Set the output name of the dnn network.
@end table
+ at subsection Examples
+
@itemize
@item
Halve the red channle of the frame with format rgb24:
@@ -9209,6 +9211,12 @@ Halve the pixel value of the frame with format gray32f:
ffmpeg -i input.jpg -vf format=grayf32,dnn_processing=model=halve_gray_float.model:input=dnn_in:output=dnn_out:dnn_backend=native -y out.native.png
@end example
+ at item
+Handle the Y channel with srcnn.pb (see @ref{sr} filter) for frame with yuv420p (planar YUV formats supported):
+ at example
+./ffmpeg -i 480p.jpg -vf format=yuv420p,scale=w=iw*2:h=ih*2,dnn_processing=dnn_backend=tensorflow:model=srcnn.pb:input=x:output=y -y srcnn.jpg
+ at end example
+
@end itemize
@section drawbox
@@ -17311,6 +17319,7 @@ Set quality level. The value @code{max} can be used to set the maximum level,
currently @code{6}.
@end table
+ at anchor{sr}
@section sr
Scale the input by applying one of the super-resolution methods based on
diff --git a/libavfilter/vf_dnn_processing.c b/libavfilter/vf_dnn_processing.c
index 4d0ee78e0b..f9458f03b2 100644
--- a/libavfilter/vf_dnn_processing.c
+++ b/libavfilter/vf_dnn_processing.c
@@ -110,6 +110,8 @@ static int query_formats(AVFilterContext *context)
static const enum AVPixelFormat pix_fmts[] = {
AV_PIX_FMT_RGB24, AV_PIX_FMT_BGR24,
AV_PIX_FMT_GRAY8, AV_PIX_FMT_GRAYF32,
+ AV_PIX_FMT_YUV420P, AV_PIX_FMT_YUV422P,
+ AV_PIX_FMT_YUV444P, AV_PIX_FMT_YUV410P, AV_PIX_FMT_YUV411P,
AV_PIX_FMT_NONE
};
AVFilterFormats *fmts_list = ff_make_format_list(pix_fmts);
@@ -163,6 +165,11 @@ static int check_modelinput_inlink(const DNNData *model_input, const AVFilterLin
}
return 0;
case AV_PIX_FMT_GRAYF32:
+ case AV_PIX_FMT_YUV420P:
+ case AV_PIX_FMT_YUV422P:
+ case AV_PIX_FMT_YUV444P:
+ case AV_PIX_FMT_YUV410P:
+ case AV_PIX_FMT_YUV411P:
if (model_input->channels != 1) {
LOG_FORMAT_CHANNEL_MISMATCH();
return AVERROR(EIO);
@@ -246,6 +253,28 @@ static int prepare_sws_context(AVFilterLink *outlink)
0, NULL, NULL, NULL);
}
return 0;
+ case AV_PIX_FMT_YUV420P:
+ case AV_PIX_FMT_YUV422P:
+ case AV_PIX_FMT_YUV444P:
+ case AV_PIX_FMT_YUV410P:
+ case AV_PIX_FMT_YUV411P:
+ av_assert0(input_dt == DNN_FLOAT);
+ av_assert0(output_dt == DNN_FLOAT);
+ ctx->sws_gray8_to_grayf32 = sws_getContext(inlink->w,
+ inlink->h,
+ AV_PIX_FMT_GRAY8,
+ inlink->w,
+ inlink->h,
+ AV_PIX_FMT_GRAYF32,
+ 0, NULL, NULL, NULL);
+ ctx->sws_grayf32_to_gray8 = sws_getContext(outlink->w,
+ outlink->h,
+ AV_PIX_FMT_GRAYF32,
+ outlink->w,
+ outlink->h,
+ AV_PIX_FMT_GRAY8,
+ 0, NULL, NULL, NULL);
+ return 0;
default:
//do nothing
break;
@@ -300,6 +329,15 @@ static int copy_from_frame_to_dnn(DnnProcessingContext *ctx, const AVFrame *fram
frame->data[0], frame->linesize[0],
bytewidth, frame->height);
return 0;
+ case AV_PIX_FMT_YUV420P:
+ case AV_PIX_FMT_YUV422P:
+ case AV_PIX_FMT_YUV444P:
+ case AV_PIX_FMT_YUV410P:
+ case AV_PIX_FMT_YUV411P:
+ sws_scale(ctx->sws_gray8_to_grayf32, (const uint8_t **)frame->data, frame->linesize,
+ 0, frame->height, (uint8_t * const*)(&dnn_input->data),
+ (const int [4]){frame->width * sizeof(float), 0, 0, 0});
+ return 0;
default:
return AVERROR(EIO);
}
@@ -341,6 +379,15 @@ static int copy_from_dnn_to_frame(DnnProcessingContext *ctx, AVFrame *frame)
dnn_output->data, bytewidth,
bytewidth, frame->height);
return 0;
+ case AV_PIX_FMT_YUV420P:
+ case AV_PIX_FMT_YUV422P:
+ case AV_PIX_FMT_YUV444P:
+ case AV_PIX_FMT_YUV410P:
+ case AV_PIX_FMT_YUV411P:
+ sws_scale(ctx->sws_grayf32_to_gray8, (const uint8_t *[4]){(const uint8_t *)dnn_output->data, 0, 0, 0},
+ (const int[4]){frame->width * sizeof(float), 0, 0, 0},
+ 0, frame->height, (uint8_t * const*)frame->data, frame->linesize);
+ return 0;
default:
return AVERROR(EIO);
}
@@ -348,6 +395,27 @@ static int copy_from_dnn_to_frame(DnnProcessingContext *ctx, AVFrame *frame)
return 0;
}
+static av_always_inline int isPlanarYUV(enum AVPixelFormat pix_fmt)
+{
+ const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(pix_fmt);
+ av_assert0(desc);
+ return !(desc->flags & AV_PIX_FMT_FLAG_RGB) && desc->nb_components == 3;
+}
+
+static int copy_uv_planes(DnnProcessingContext *ctx, AVFrame *out, const AVFrame *in)
+{
+ const AVPixFmtDescriptor *desc = av_pix_fmt_desc_get(in->format);
+ int uv_height = AV_CEIL_RSHIFT(in->height, desc->log2_chroma_h);
+ for (int i = 1; i < 3; ++i) {
+ int bytewidth = av_image_get_linesize(in->format, in->width, i);
+ av_image_copy_plane(out->data[i], out->linesize[i],
+ in->data[i], in->linesize[i],
+ bytewidth, uv_height);
+ }
+
+ return 0;
+}
+
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
{
AVFilterContext *context = inlink->dst;
@@ -373,6 +441,10 @@ static int filter_frame(AVFilterLink *inlink, AVFrame *in)
av_frame_copy_props(out, in);
copy_from_dnn_to_frame(ctx, out);
+
+ if (isPlanarYUV(in->format))
+ copy_uv_planes(ctx, out, in);
+
av_frame_free(&in);
return ff_filter_frame(outlink, out);
}
More information about the ffmpeg-cvslog
mailing list