[FFmpeg-devel] [PATCH] yuv pixel formats support in openjpeg decoder + 10bit support
Dave Rice
daverice at mac.com
Sat Nov 12 14:07:42 CET 2011
On Nov 12, 2011, at 6:39 AM, Alex Zhukov <zhukov.alex at gmail.com> wrote:
> re code cleanups: see the new patch
> re curly braces: i'd keep them to make further patches less in size
>
> re rgb48, needs a little bit of context
> there is no field in j2k (i know of) that tells you what color space
> is used. i am guessing even if there was one people would rarely use
> it.
AFAIK this is right, the container must communicate the codec as in the COLR chunk or SMPTE 422m. There's an implementation of this in the gstreamer bad plugins.
I've been traveling this weekend but will upload more j2k samples late tonight.
Dave Rice
> so most of the code in check_image_attributes() is guesswork that
> basically does the following:
> - if component sizes are not the same its more likely to be YUV (ratio 1/2)
> - if component sizes ARE the same and precision of component is 8 bit its RGB24
>
> what do we do with higher precisions is, imho, a matter of taste,
> 16bpp equal component size j2k can either be YUV444P16 or RGB48.
>
> returning PIX_FMT_YUV444P16 i made an assumption that it is more
> likely to be YUV444P16, mostly because i encounter such files more
> often.
> tell me if you have a different experience.
>
>
> On Sat, Nov 12, 2011 at 1:13 AM, Jean First <jeanfirst at gmail.com> wrote:
>> On Fri Nov 11 2011 17:09:01 GMT+0100 (CET), Alex Zhukov wrote:
>>>
>>> rgb24 works
>>>
>>> On Fri, Nov 11, 2011 at 7:13 AM, Alex Zhukov<zhukov.alex at gmail.com>
>>> wrote:
>>>>
>>>> good point
>>>> will fix
>>
>> [...]
>>
>>> From 29c626cbb5b56d7362ef34ad9751c2ee6718a997 Mon Sep 17 00:00:00 2001
>>> From: Alex Zhukov <zhukov.alex at gmail.com>
>>> Date: Thu, 10 Nov 2011 14:08:24 -0800
>>> Subject: [PATCH] yuv pixel formats support in openjpeg decoder + 10bit
>>> support
>>>
>>> Signed-off-by: Alex Zhukov <zhukov.alex at gmail.com>
>>> ---
>>> libavcodec/libopenjpeg.c | 161
>>> ++++++++++++++++++++++++++++++++++++----------
>>> 1 files changed, 126 insertions(+), 35 deletions(-)
>>>
>>> diff --git a/libavcodec/libopenjpeg.c b/libavcodec/libopenjpeg.c
>>> index 42c71e3..0d48ad4 100644
>>> --- a/libavcodec/libopenjpeg.c
>>> +++ b/libavcodec/libopenjpeg.c
>>> @@ -25,6 +25,7 @@
>>> */
>>>
>>> #include "libavutil/imgutils.h"
>>> +#include "libavutil/pixfmt.h"
>>> #include "avcodec.h"
>>> #include "libavutil/intreadwrite.h"
>>> #include "thread.h"
>>> @@ -39,14 +40,105 @@ typedef struct {
>>> AVFrame image;
>>> } LibOpenJPEGContext;
>>>
>>> -static int check_image_attributes(opj_image_t *image)
>>> +static enum PixelFormat check_image_attributes(AVCodecContext *avctx,
>>> opj_image_t *image)
>>> {
>>> - return image->comps[0].dx == image->comps[1].dx &&
>>> - image->comps[1].dx == image->comps[2].dx &&
>>> - image->comps[0].dy == image->comps[1].dy &&
>>> - image->comps[1].dy == image->comps[2].dy &&
>>> - image->comps[0].prec == image->comps[1].prec &&
>>> - image->comps[1].prec == image->comps[2].prec;
>>> + opj_image_comp_t c0 = image->comps[0];
>>> + opj_image_comp_t c1 = image->comps[1];
>>> + opj_image_comp_t c2 = image->comps[2];
>>> + int compRatio = c0.dx << 15 | c0.dy << 12;
>>> + compRatio |= c1.dx << 9 | c1.dy << 6;
>>> + compRatio |= c2.dx << 3 | c2.dy << 0;
>>> +
>>> + switch(compRatio) {
>>> + case 0111111: goto libopenjpeg_yuv444_rgb;
>>> + case 0112121: goto libopenjpeg_yuv422;
>>> + case 0112222: goto libopenjpeg_yuv420;
>>> + default: return PIX_FMT_RGB24;
>>> + }
>>
>> What about rgb48 ? Can you test it ?
>>
>>> +
>>> +libopenjpeg_yuv420:
>>> + switch (c0.prec) {
>>> + case 8: return PIX_FMT_YUV420P;
>>> + case 9: return PIX_FMT_YUV420P9;
>>> + case 10: return PIX_FMT_YUV420P10;
>>> + case 16: return PIX_FMT_YUV420P16;
>>> + }
>>> +
>>> +libopenjpeg_yuv422:
>>> + switch (c0.prec) {
>>> + case 8: return PIX_FMT_YUV422P;
>>> + case 9: return PIX_FMT_YUV422P9;
>>> + case 10: return PIX_FMT_YUV422P10;
>>> + case 16: return PIX_FMT_YUV422P16;
>>> + }
>>> +
>>> +libopenjpeg_yuv444_rgb:
>>> + switch (c0.prec) {
>>> + case 8: return PIX_FMT_RGB24;
>>> + case 9: return PIX_FMT_YUV444P9;
>>> + case 10: return PIX_FMT_YUV444P10;
>>> + case 16: return PIX_FMT_YUV444P16;
>>> + }
>>> + return PIX_FMT_RGB24;
>>> +}
>>> +
>>> +static inline int libopenjpeg_ispacked(enum PixelFormat pix_fmt) {
>>> + int i, component_plane;
>>> + component_plane = av_pix_fmt_descriptors[pix_fmt].comp[0].plane;
>>> + for(i=1; i < av_pix_fmt_descriptors[pix_fmt].nb_components;i++) {
>>> + if (component_plane !=
>>> av_pix_fmt_descriptors[pix_fmt].comp[i].plane)
>>> + return 0;
>>> + }
>>> + return 1;
>>> +}
>>> +
>>> +static inline void libopenjpeg_copy_to_packed8(AVFrame *picture,
>>> opj_image_t *image) {
>>> + uint8_t *img_ptr;
>>> + int index, x, y, c;
>>> + for(y = 0; y < picture->height; y++) {
>>> + index = y*picture->width;
>>> + img_ptr = picture->data[0] + y*picture->linesize[0];
>>> + for(x = 0; x < picture->width; x++, index++) {
>>> + for(c = 0; c < image->numcomps; c++) {
>>> + *img_ptr++ = image->comps[c].data[index];
>>> + }
>>> + }
>>> + }
>>> +}
>>> +
>>> +static inline void libopenjpeg_copyto8(AVFrame *picture, opj_image_t
>>> *image) {
>>> + int *comp_data;
>>> + uint8_t *img_ptr;
>>> + int index, x, y;
>>> +
>>> + for(index = 0; index < image->numcomps; index++) {
>>> + comp_data = image->comps[index].data;
>>> + img_ptr = picture->data[index];
>>> + for(y = 0; y < image->comps[index].h; y++) {
>>> + for(x = 0; x < image->comps[index].w; x++) {
>>> + *img_ptr = (uint8_t) *comp_data;
>>> + img_ptr++;
>>> + comp_data++;
>>> + }
>>> + }
>>> + }
>>> +}
>>> +
>>> +static inline void libopenjpeg_copyto16(AVFrame *picture, opj_image_t
>>> *image) {
>>> + int *comp_data;
>>> + uint16_t *img_ptr;
>>> + int index, x, y;
>>> + for(index = 0; index < image->numcomps; index++) {
>>> + comp_data = image->comps[index].data;
>>> + img_ptr = (uint16_t*) picture->data[index];
>>> + for(y = 0; y < image->comps[index].h; y++) {
>>> + for(x = 0; x < image->comps[index].w; x++) {
>>> + *img_ptr = *comp_data;
>>> + img_ptr++;
>>> + comp_data++;
>>> + }
>>> + }
>>> + }
>>> }
>>>
>>> static av_cold int libopenjpeg_decode_init(AVCodecContext *avctx)
>>> @@ -78,10 +170,9 @@ static int libopenjpeg_decode_frame(AVCodecContext
>>> *avctx,
>>> opj_dinfo_t *dec;
>>> opj_cio_t *stream;
>>> opj_image_t *image;
>>> - int width, height, has_alpha = 0, ret = -1;
>>> - int x, y, index;
>>> - uint8_t *img_ptr;
>>> - int adjust[4];
>>> + int width, height, ret = -1;
>>> + int pixel_size = 0;
>>> + int ispacked = 0;
>>>
>>> *data_size = 0;
>>>
>>> @@ -115,7 +206,7 @@ static int libopenjpeg_decode_frame(AVCodecContext
>>> *avctx,
>>> }
>>>
>>> // Decode the header only
>>> - image = opj_decode_with_info(dec, stream, NULL);
>>> + image = opj_decode(dec, stream);
>>
>> this is not needed. opj_decode calls opj_decode_with_info !
>>
>>> opj_cio_close(stream);
>>> if(!image) {
>>> av_log(avctx, AV_LOG_ERROR, "Error decoding codestream.\n");
>>> @@ -134,15 +225,9 @@ static int libopenjpeg_decode_frame(AVCodecContext
>>> *avctx,
>>> {
>>> case 1: avctx->pix_fmt = PIX_FMT_GRAY8;
>>> break;
>>> - case 3: if(check_image_attributes(image)) {
>>> - avctx->pix_fmt = PIX_FMT_RGB24;
>>> - } else {
>>> - avctx->pix_fmt = PIX_FMT_GRAY8;
>>> - av_log(avctx, AV_LOG_ERROR, "Only first component
>>> will be used.\n");
>>> - }
>>> + case 3: avctx->pix_fmt = check_image_attributes(avctx, image);
>>> break;
>>> - case 4: has_alpha = 1;
>>> - avctx->pix_fmt = PIX_FMT_RGBA;
>>> + case 4: avctx->pix_fmt = PIX_FMT_RGBA;
>>> break;
>>> default: av_log(avctx, AV_LOG_ERROR, "%d components
>>> unsupported.\n", image->numcomps);
>>> goto done;
>>> @@ -170,25 +255,31 @@ static int libopenjpeg_decode_frame(AVCodecContext
>>> *avctx,
>>> }
>>>
>>> // Decode the codestream
>>> - image = opj_decode_with_info(dec, stream, NULL);
>>> + image = opj_decode(dec, stream);
>>
>> see above.
>>
>>> opj_cio_close(stream);
>>>
>>> - for(x = 0; x < image->numcomps; x++) {
>>> - adjust[x] = FFMAX(image->comps[x].prec - 8, 0);
>>> - }
>>> + pixel_size =
>>> av_pix_fmt_descriptors[avctx->pix_fmt].comp[0].step_minus1+1;
>>> + ispacked = libopenjpeg_ispacked(avctx->pix_fmt);
>>>
>>> - for(y = 0; y < avctx->height; y++) {
>>> - index = y*avctx->width;
>>> - img_ptr = picture->data[0] + y*picture->linesize[0];
>>> - for(x = 0; x < avctx->width; x++, index++) {
>>> - *img_ptr++ = image->comps[0].data[index] >> adjust[0];
>>> - if(image->numcomps > 2 && check_image_attributes(image)) {
>>> - *img_ptr++ = image->comps[1].data[index] >> adjust[1];
>>> - *img_ptr++ = image->comps[2].data[index] >> adjust[2];
>>> - if(has_alpha)
>>> - *img_ptr++ = image->comps[3].data[index] >>
>>> adjust[3];
>>> + switch(pixel_size) {
>>> + case 1:
>>> + if (ispacked) {
>>> + libopenjpeg_copy_to_packed8(picture, image);
>>> + } else {
>>> + libopenjpeg_copyto8(picture, image);
>>> }
>>> - }
>>> + break;
>>> + case 2:
>>> + libopenjpeg_copyto16(picture, image);
>>> + break;
>>> + case 3:
>>> + if (ispacked) {
>>> + libopenjpeg_copy_to_packed8(picture, image);
>>> + }
>>> + break;
>>> + default:
>>> + av_log(avctx, AV_LOG_ERROR, "unsupported pixel size %d\n",
>>> pixel_size);
>>> + goto done;
>>> }
>>>
>>> *output = ctx->image;
>>> --
>>> 1.7.4.4
>>>
>>
>> _______________________________________________
>> ffmpeg-devel mailing list
>> ffmpeg-devel at ffmpeg.org
>> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>>
> <0001-yuv-pixel-formats-support-in-openjpeg-decoder-10bit-.patch>
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
More information about the ffmpeg-devel
mailing list