Intel® Integrated Performance Primitives
Deliberate problems developing high-performance vision, signal, security, and storage applications.
6704 Discussions

ippiResizeSql_Pixel16u_C1R returning -23 (ippStsResizeFactorErr)

Aris_Basic
New Contributor I
380 Views
im trying tor esize a 16bit image but im always getting back -23 error, and xFactor and yFactor are definetly not 0 or close to 0 (in this case they are close to 32)
Im using latest ipp 7.0 with IPP_INTER_SUPER as interpolation method
(pretty much i created a copy of CIppImage::Resize method so it can handle 16bit data
and replaced all ippiResizeSql_Pixel_8u calls with 16u calls
(currently CIppImage::Resize can handle 8bit images only)
Could it be issue with ippiResizeGetBufSize ?? (no parameters for bit size of the data it is going to be used for)
0 Kudos
4 Replies
Aris_Basic
New Contributor I
381 Views
Ok after reading docs for depricated function ippiResize.
i figgued out that its is the issue with interplation IPPI_INTER_SUPER that is sued by default in CIppImage::Resize
also Depricated function returns error -22 (Invalid Interpolation Method) when trying to enlarge image and
using IPP_INTER_SUPER while ResizeSqlPixel returns -23 (bad Resize Factor) i think this is a bug in this function
Thanks
0 Kudos
Vladimir_Dudnik
Employee
380 Views
Hi Aris,

I did not quite well get your point. What is the bug? Could you please specify actual parameters you pass to the function when it return error code?

Regards,
Vladimir
0 Kudos
Aris_Basic
New Contributor I
381 Views
wrong error code
if i want to Resize to bgger image and pass Super sampling as interpolation i get error code Bad Resize Factor -23(0 or less then 0), instead of bad interpolation -22 as with old function.
also i think that need to be noted in documentation
(maybe UIC sample code CIppImage should use LINEAR as default interpolation in Resize method instead of SUPER to remove confusion)
0 Kudos
Vladimir_Dudnik
Employee
381 Views
AccordingIPP documentation Super Sampling interpolation can only be used to reduce image resolution. So when one try to use it to enlarge image something is definetely wrong either interpolation mode or resize factors. I agree the difference between new anddeprecated implementation of resize operation should be clearly stated in documentation.

Also thanks for pointing to UIC image sample class, we will correct that.

Regards,
Vladimir
0 Kudos
Reply