[FFmpeg-devel] [PATCH] libopenjpeg wrapper for jpeg2k decoding
Michael Niedermayer
michaelni
Wed Feb 4 09:42:52 CET 2009
On Wed, Feb 04, 2009 at 10:33:52AM +0530, Jai Menon wrote:
> Hi,
>
> On Tue, Feb 3, 2009 at 5:51 PM, Michael Niedermayer <michaelni at gmx.at> wrote:
> > On Tue, Feb 03, 2009 at 11:31:23AM +0530, Jai Menon wrote:
> >> Hi,
> >>
> >> On Tue, Feb 3, 2009 at 12:58 AM, Michael Niedermayer <michaelni at gmx.at> wrote:
> >> > On Mon, Feb 02, 2009 at 04:48:17PM +0530, Jai Menon wrote:
> >> >> Hi,
> >> >>
> >> >> 2009/1/29 Michael Niedermayer <michaelni at gmx.at>:
> >> >> > On Thu, Jan 29, 2009 at 11:35:31AM +0530, Jai Menon wrote:
> >> >> >> Hi,
> >> >> >>
> >> >> >> On Thu, Jan 29, 2009 at 4:27 AM, Michael Niedermayer <michaelni at gmx.at> wrote:
> >> >> >> > On Wed, Jan 28, 2009 at 02:23:46PM +0530, Jai Menon wrote:
> >> >>
> >> >> [...]
> >> >>
> >> >> > dec is written to twice with no read in between, this applies to more vars
> >> >>
> >> >> fixed.
> >> >>
> >> >> [...]
> >> >>
> >> >> > sec hole when width/height changes
> >> >>
> >> >> fixed (i think).
> >> >>
> >> >> i also reworked most of the code to use get/release_buffer
> >> >> revised patch attached.
> >> > [...]
> >> >> +static int libopenjpeg_decode_frame(AVCodecContext *avctx,
> >> >> + void *data, int *data_size,
> >> >> + const uint8_t *buf, int buf_size)
> >> >> +{
> >> >> + LibOpenJPEGContext *ctx = avctx->priv_data;
> >> >> + AVFrame *picture = &ctx->image, *output = data;
> >> >> + opj_dinfo_t *dec;
> >> >
> >> >> + opj_cio_t *stream = NULL;
> >> >
> >> > written twice before read
> >>
> >> this and another one fixed.
> >>
> >> >> + opj_image_t *image = NULL;
> >> >> + int width, height, has_alpha = 0, ret = -1;
> >> >> + int x, y, index;
> >> >> + uint8_t *img_ptr;
> >> >> + float scale;
> >> >> +
> >> >> + *data_size = 0;
> >> >> +
> >> >> + // Check if input is a raw jpeg2k codestream or in jp2 wrapping
> >> >> + if((AV_RB32(buf) == 12) &&
> >> >> + (AV_RB32(buf + 4) == JP2_SIG_TYPE) &&
> >> >> + (AV_RB32(buf + 8) == JP2_SIG_VALUE)) {
> >> >> + dec = opj_create_decompress(CODEC_JP2);
> >> >
> >> >> + }
> >> >> + else dec = opj_create_decompress(CODEC_J2K);
> >> >
> >> > diego likely wont like this kind of } placement style
> >>
> >> changed.
> >>
> >> > [...]
> >> >> + scale = 255.0f / (float)((1 << image->comps[0].prec) - 1);
> >> >> + for(y = 0; y < height; y++) {
> >> >> + index = y*width;
> >> >> + img_ptr = picture->data[0] + y*picture->linesize[0];
> >> >> + for(x = 0; x < width; x++, index++) {
> >> >> + *img_ptr++ = image->comps[0].data[index] * scale;
> >> >> + if(image->numcomps > 2 && check_image_attributes(image)) {
> >> >> + *img_ptr++ = image->comps[1].data[index] * scale;
> >> >> + *img_ptr++ = image->comps[2].data[index] * scale;
> >> >> + if(has_alpha)
> >> >> + *img_ptr++ = image->comps[3].data[index] * scale;
> >> >> + }
> >> >> + }
> >> >> + }
> >> >
> >> > lrintf()
> >>
> >> hmm...i made this generic and modified it to use integer arithmetic. i
> >> think its better.
> >> revised patch attached.
> >>
> >
> > [...]
> >
> >> + if(image->comps[x].prec > 8)
> >> + adjust[x] = image->comps[x].prec - 8;
> >> + else adjust[x] = 0;
> >
> > there should be a {} between the if/else, it doesnt add any extra lines
> > but allows something to be inserted without changing the if/else lines, that
> > is it means future patches would be simpler
> >
> > but then i realize now this could be written simpler with FFMAX()
>
> changed.
>
> >> + }
> >> +
> >> + for(y = 0; y < height; y++) {
> >> + index = y*width;
> >> + img_ptr = picture->data[0] + y*picture->linesize[0];
> >> + for(x = 0; x < width; x++, index++) {
> >> + *img_ptr++ = (image->comps[0].data[index] >> adjust[0]) +
> >> + (image->comps[0].data[index] >> (adjust[0] - 1) & 1);
> >
> > if adjust[x] == 0 -> A >> (-1) which is undefined
> >
> > also this looks like (data+(1<<(adjust-1)))>>adjust
>
> made it much simpler since i didn't see any difference in the test images.
[...]
> + // Check if input is a raw jpeg2k codestream or in jp2 wrapping
> + if((AV_RB32(buf) == 12) &&
> + (AV_RB32(buf + 4) == JP2_SIG_TYPE) &&
> + (AV_RB32(buf + 8) == JP2_SIG_VALUE)) {
> + dec = opj_create_decompress(CODEC_JP2);
> + } else {
> + dec = opj_create_decompress(CODEC_J2K);
one of these is wrongly indented
> + }
> +
> + if(!dec) {
> + av_log(avctx, AV_LOG_ERROR, "Error initializing decoder.\n");
> + return -1;
> + }
> + opj_set_event_mgr((opj_common_ptr)dec, NULL, NULL);
> + // Tie decoder with decoding parameters
> + opj_setup_decoder(dec, &ctx->dec_params);
> + stream = opj_cio_open((opj_common_ptr)dec, buf, buf_size);
> + if(!stream) {
> + av_log(avctx, AV_LOG_ERROR, "Codestream could not be opened for reading.\n");
> + opj_destroy_decompress(dec);
> + return -1;
> + }
> + // Decode the codestream
> + image = opj_decode_with_info(dec, stream, NULL);
i think an empty line before the // comments would improve readabilty
[...]
> +AVCodec libopenjpeg_decoder = {
> + "libopenjpeg",
> + CODEC_TYPE_VIDEO,
> + CODEC_ID_JPEG2000,
> + sizeof(LibOpenJPEGContext),
> + libopenjpeg_decode_init,
> + NULL,
> + libopenjpeg_decode_close,
> + libopenjpeg_decode_frame,
> + CODEC_CAP_DELAY,
looks wrong
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
I have never wished to cater to the crowd; for what I know they do not
approve, and what they approve I do not know. -- Epicurus
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20090204/f10dac0f/attachment.pgp>
More information about the ffmpeg-devel
mailing list