Evaluates variances and means of four 8x8 blocks of difference between two 16x16 blocks.
Syntax
IppStatus ippiVarMeanDiff16x16_8u32s_C1R(const Ipp8u* pSrc, Ipp32s srcStep, const Ipp8u* pRef, Ipp32s srcRefStep, Ipp32s* pSrcSum, Ipp32s* pVar, Ipp32s* pMean, Ipp32s mcType);
Parameters
pSrc |
Pointer to the current block 16x16. |
srcStep |
Current block step. |
pRef |
Pointer to the reference block 16x16. |
srcRefStep |
Reference block step. |
pSrcSum |
Pointer to the sum of pixel values for the current block. |
pVar |
Pointer to the variance. |
pMean |
Pointer to the mean value. |
mcType |
MC type IPPVC_MC_APX. |
Description
This function is declared in the ippvc.h header file. The function ippiVarMeanDiff16x16_8u32s_C1R evaluates the means and variances of four 8x8 blocks of difference between two 16x16 blocks.
As the function is repeatedly called for the same current block, the parameter pSrcSum is used to avoid excessive computation. This parameter should point to an array of four 8x8 blocks comprising the current block. The parameters pVar, pMean should point to arrays of four elements, since the values of variance and mean are calculated separately for each 8x8 block.
Mean is evaluated by the following formula:
where Sum is the sum of pixel values for the current block taken from pSrcSum parameter.
Variance is evaluated by the following formula:
This function is used in the MPEG-2 encoder included into Intel IPP Samples. See introduction to this section.
Return Values
ippStsNoErr |
Indicates no error. |
ippStsNullPtrErr |
Indicates an error when at least one input pointer is NULL. |