- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Here is the issue with chroma decoding.
I feed to decoder the file:
And get decoded:
The picture is coded as JC_YCBCR with subsampling JS_422
If I decode this picture into JC_GRAY - I get goot quality grayscale picture.
But if I require to decode this picture into the colorspace and subsampling
as it is coded internally: JC_YCBCR with subsampling JS_422
I see wrong image.
How to fix it?
I feed to decoder the file:
And get decoded:
The picture is coded as JC_YCBCR with subsampling JS_422
If I decode this picture into JC_GRAY - I get goot quality grayscale picture.
But if I require to decode this picture into the colorspace and subsampling
as it is coded internally: JC_YCBCR with subsampling JS_422
I see wrong image.
How to fix it?
Link Copied
3 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello,
How do you set the output video format? some samle code may be helpful here:
I see from the MJPEG sample code, it needs to set as YUY2:
How do you set the output video format? some samle code may be helpful here:
I see from the MJPEG sample code, it needs to set as YUY2:
if(JS_422 == sampling && JC_YCBCR == color)
{
m_colorFormat = YUY2;
m_frameStep = m_frameDims.width*2;
}
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
[bash]My code is there^
jerr = m_decoder.ReadHeader(&in_width, &in_height, &in_channels, &in_color, &in_sampling, &in_precision);
// Allocate buffer here jerr = m_decoder.SetDestination(p42xP4, n42xStepP4, dim4xxY, in_channels, in_color, in_sampling, 8, scale); jerr = m_decoder.ReadData();[/bash]
In this code colorspace and sampling are exactly as coded in compressed JPEG image.
I do so, to avoid redundant color convertions inside of decoder.
Why should I change it?
In IPP 6.1 everithing worked fine.
Why it has been changed?
jerr = m_decoder.ReadHeader(&in_width, &in_height, &in_channels, &in_color, &in_sampling, &in_precision);
// Allocate buffer here jerr = m_decoder.SetDestination(p42xP4, n42xStepP4, dim4xxY, in_channels, in_color, in_sampling, 8, scale); jerr = m_decoder.ReadData();[/bash]
In this code colorspace and sampling are exactly as coded in compressed JPEG image.
I do so, to avoid redundant color convertions inside of decoder.
Why should I change it?
In IPP 6.1 everithing worked fine.
Why it has been changed?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello!
Problem still exists with ipp-samples ver. 7.0.7. with JPEG 4:2:2 YCbCr. Is it possible to avoid YUY2 in output and get YUV422?
Possibly i have to change 2928 line in jpegdec.cpp from
status = ippiYCbCr422_8u_P3C2R(pSrc8u,srcStep,pDst8u,dstStep,roi);
to something else, or add some other conversion?
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page