[x265] [PATCH] calculate SSIM for each Row after deblock, sao
Steve Borho
steve at borho.org
Tue Oct 8 19:31:36 CEST 2013
On Tue, Oct 8, 2013 at 12:13 PM, Aarthi Priya Thirumalai <
aarthi at multicorewareinc.com> wrote:
>
>
>
> On Tue, Oct 8, 2013 at 10:29 PM, Steve Borho <steve at borho.org> wrote:
>
>>
>>
>>
>> On Tue, Oct 8, 2013 at 6:43 AM, Aarthi Thirumalai <
>> aarthi at multicorewareinc.com> wrote:
>>
>>> # HG changeset patch
>>> # User Aarthi Thirumalai
>>> # Date 1381232542 -19800
>>> # Tue Oct 08 17:12:22 2013 +0530
>>> # Node ID 3ba555e98be2c0641d92d14155c6fb303c0003d9
>>> # Parent 47286888d5a293234340153112810ce2a4f81546
>>> calculate SSIM for each Row after deblock, sao
>>>
>>
>> I've queued this and the other patch after fixing more nits
>>
>>
>>>
>>> diff -r 47286888d5a2 -r 3ba555e98be2 source/Lib/TLibEncoder/TEncTop.cpp
>>> --- a/source/Lib/TLibEncoder/TEncTop.cpp Tue Oct 08 17:06:57 2013
>>> +0530
>>> +++ b/source/Lib/TLibEncoder/TEncTop.cpp Tue Oct 08 17:12:22 2013
>>> +0530
>>> @@ -513,7 +513,7 @@
>>> int width = recon->getWidth() - getPad(0);
>>> int height = recon->getHeight() - getPad(1);
>>> int size = width * height;
>>> -
>>> + double ssim = 0;
>>> UInt64 ssdY = computeSSD(orig->getLumaAddr(), recon->getLumaAddr(),
>>> stride, width, height);
>>>
>>> height >>= 1;
>>> @@ -604,7 +604,14 @@
>>> {
>>> m_analyzeB.addResult(psnrY, psnrU, psnrV, (double)bits);
>>> }
>>> -
>>> + if (param.bEnableSsim)
>>> + {
>>> + if(pic->getSlice()->m_ssimCnt > 0)
>>>
>> white-space
>>
>>> + {
>>> + ssim += pic->getSlice()->m_ssim /
>>> pic->getSlice()->m_ssimCnt;
>>>
>> I declared double ssim = pic->.. here since `ssim' is used nowhere else
>>
>
> I had added ssim var to print out along its value along with psnr values
> at the end of this fuction. guess I haven't committed that change. ll do
> it over this, once its pushed.
>
>> + m_globalSsim += ssim;
>>> + }
>>> + }
>>> if (param.logLevel >= X265_LOG_DEBUG)
>>> {
>>> char c = (slice->isIntra() ? 'I' : slice->isInterP() ? 'P' :
>>> 'B');
>>> diff -r 47286888d5a2 -r 3ba555e98be2 source/encoder/frameencoder.cpp
>>> --- a/source/encoder/frameencoder.cpp Tue Oct 08 17:06:57 2013 +0530
>>> +++ b/source/encoder/frameencoder.cpp Tue Oct 08 17:12:22 2013 +0530
>>> @@ -92,7 +92,7 @@
>>> }
>>>
>>> m_frameFilter.destroy();
>>> -
>>> + X265_FREE(ssimBuf);
>>>
>>
>> renamed ssimBuf to m_ssimBuf to follow class member naming convention
>>
>>
>>> // wait for worker thread to exit
>>> stop();
>>> }
>>> @@ -111,6 +111,9 @@
>>> m_rows[i].create(top);
>>> }
>>>
>>> + if (m_cfg->param.bEnableSsim)
>>> + CHECKED_MALLOC(ssimBuf, ssim_t, 8 * (m_cfg->param.sourceWidth /
>>> 4 + 3));
>>>
>>
>> It's actually *safer* to not use CHECKED_MALLOC here. You're already
>> checking for NULL pointer below before using the buffer, so all the checked
>> malloc does here is skip the rest of the class initialization, including
>> the part that starts the worker thread.. so instead of a bad malloc causing
>> no SSIM results to be displayed, it would cause the encoder to deadlock.
>>
>
> aah yes. got it.
>
>>
>>
>>> +
>>> // NOTE: 2 times of numRows because both Encoder and Filter in same
>>> queue
>>> if (!WaveFront::init(m_numRows * 2))
>>> {
>>> @@ -168,6 +171,8 @@
>>> assert(0);
>>> }
>>> start();
>>> +fail:
>>> + return;
>>> }
>>>
>>> int FrameEncoder::getStreamHeaders(NALUnitEBSP **nalunits)
>>> @@ -535,6 +540,31 @@
>>> slice->setSaoEnabledFlag((saoParam->bSaoFlag[0] == 1) ? true :
>>> false);
>>> }
>>>
>>> + /*Compute SSIM if enabled*/
>>>
>>
>> added spaces within /* and */
>>
>> + if (m_cfg->param.bEnableSsim && ssimBuf)
>>> + {
>>> + pixel *rec = (pixel*)m_pic->getPicYuvRec()->getLumaAddr();
>>> + pixel *org = (pixel*)m_pic->getPicYuvOrg()->getLumaAddr();
>>> + int stride1 = m_pic->getPicYuvOrg()->getStride();
>>> + int stride2 = m_pic->getPicYuvRec()->getStride();
>>> + for (int row = 0; row < m_numRows; row++)
>>> + {
>>> + int bEnd = ((row + 1) == (this->m_numRows - 1));
>>> + int bStart = (row == 0);
>>> + int minPixY = row * 64 - 4 * !bStart;
>>> + int maxPixY = (row + 1) * 64 - 4 * !bEnd;
>>> + int ssim_cnt;
>>> + x265_emms();
>>> +
>>> + /* SSIM is done for each row in blocks of 4x4 . The First
>>> blocks are offset by 2 pixels to the right
>>> + * to avoid alignment of ssim blocks with DCT blocks. */
>>> + minPixY += bStart ? 2 : -6;
>>> + slice->m_ssim += calculateSSIM(rec + 2 + minPixY * stride1,
>>> stride1, org + 2 + minPixY * stride2, stride2,
>>> + m_cfg->param.sourceWidth - 2, maxPixY - minPixY,
>>> ssimBuf, &ssim_cnt);
>>>
>>
>> lined up the second line of arguments with the first
>>
>> + slice->m_ssimCnt += ssim_cnt;
>>> + }
>>> + }
>>> +
>>> entropyCoder->setBitstream(NULL);
>>>
>>> // Reconstruction slice
>>> @@ -681,6 +711,39 @@
>>> delete bitstreamRedirect;
>>> }
>>>
>>> +/* Function to calculate SSIM for each row */
>>> +float FrameEncoder::calculateSSIM(pixel *pix1, intptr_t stride1, pixel
>>> *pix2, intptr_t stride2, int width, int height, void *buf, int *cnt)
>>> +{
>>> + int z = 0;
>>> + float ssim = 0.0;
>>> + ssim_t(*sum0)[4] = (ssim_t(*)[4])buf;
>>> + ssim_t(*sum1)[4] = sum0 + (width >> 2) + 3;
>>> + width >>= 2;
>>> + height >>= 2;
>>> +
>>> + for (int y = 1; y < height; y++)
>>> + {
>>> + for (; z <= y; z++)
>>> + {
>>> + void* swap = sum0;
>>> + sum0 = sum1;
>>> + sum1 = (ssim_t(*)[4])swap;
>>> + for (int x = 0; x < width; x += 2)
>>> + {
>>> + primitives.ssim_4x4x2_core(&pix1[(4 * x + (z *
>>> stride1))], stride1, &pix2[(4 * x + (z * stride2))], stride2, &sum0[x]);
>>> + }
>>> + }
>>> +
>>> + for (int x = 0; x < width - 1; x += 4)
>>> + {
>>> + ssim += primitives.ssim_end_4(sum0 + x, sum1 + x,
>>> X265_MIN(4, width - x - 1));
>>> + }
>>> + }
>>> +
>>> + *cnt = (height - 1) * (width - 1);
>>> + return ssim;
>>> +}
>>> +
>>> void FrameEncoder::encodeSlice(TComOutputBitstream* substreams)
>>> {
>>> // choose entropy coder
>>> diff -r 47286888d5a2 -r 3ba555e98be2 source/encoder/frameencoder.h
>>> --- a/source/encoder/frameencoder.h Tue Oct 08 17:06:57 2013 +0530
>>> +++ b/source/encoder/frameencoder.h Tue Oct 08 17:12:22 2013 +0530
>>> @@ -145,6 +145,9 @@
>>> /* called by compressFrame to perform wave-front compression
>>> analysis */
>>> void compressCTURows();
>>>
>>> + /* called by compressFrame to calculate SSIM for each row . */
>>>
>>
>> fixed white-space in the comment
>>
>
> any better way to catch such white space issues before commit? uncrustify
> doesnt seem to fix all of this.
>
No, it just takes an eye practiced in the art of bikeshedding.
--
Steve Borho
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20131008/3953cf74/attachment-0001.html>
More information about the x265-devel
mailing list