Problem with ippiResizeSqrPixel

[postaquestion]

I use

st = ippiResizeSqrPixel_32f_C1R(m_storage, sz, m_wstep, srcroi, m_storage, m_wstep, dstroi, 1.0, 1.0, 0.0, 0.0, IPPI_INTER_LINEAR, pBuffer);

since the xfactor, yfactor are 1.0, the image should not be changed. actually it is not.

but if I use IPPI_INTER_NN, or CUBIC, LANCSOZ, the results are correct. Do you have any idea?

The image size is 10x10.

Thanks

[Ying_S_Intel]

What version of Intel IPP did you use? Can you send us your test file along with image file so we can verify it?

Thanks,
Ying S
Intel IPP Support

[postaquestion]

Iused version 5.2 beta.

the image is quite simple

1 2 3 4 5 6 7 8 9 10
11 12 13 14 15...
..
..99

it is a sequence from 1 to 99.

Ifound other problems.

The resizing function can't be in-place,I mean the input buffer and output buffer are the same. Doyou use temporary buffer during the computation?

The cubic interpolation generated a runtime error if the oringinal image size is 3680x4752, the rescaled image is 500x4752. the xfactor is 0.135869 and yfactor is 1.0. Canyou test on that image?Or isthe xfactor not correct?

Thanks a lot.

[Vladimir_Dudnik]

Hello,

regarding runtime error, we confirm that there is an issue in IPP 5.2 beta optimized code. We will fix this issue in IPP 5.2 final release which is targeted sometime in Q2 2007

Regards,
Vladimir

[Intel_Software_Netw1]

Hello,

This issue has been fixed in IPP 5.2 Gold.

Regards,

Gina B.
Intel Software Network Support
http://www.intel.com/software
email: ISN.support@intel.com

Intel is a registered trademark of Intel Corporation or its subsidiaries in the United States and other countries.

*Other names and brands may be claimed as the property of others.