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

Wiener filter

szakatszozo
Débutant
773 Visites
Hi everybody,
I would like to restore a blurred image by constant object motion, useing the Wiener filter.
I can calculate the motion of the object in pixels in the captured image and also the angle of the motion. Is it possible to use the ippiFilterWiener to remove the blurr or that is only made for noise reduction?
What other possibilities I have with IPP functions?
Thank you in advance for the help,
zozo'
0 Compliments
2 Réponses
Vladimir_Dudnik
Employé
773 Visites
Hi,
Yes, Wiener is for denoise operation. Unsharp filtering can help

/// another way, shorter. just one operation with unsharp mask

Ipp32f kernel[] = { -0.1667f, -0.6667f, -0.1667f, -0.6667f, 4.3333f, -0.6667f, -0.1667f, -0.6667f, -0.1667f };

IppiSize kernelSize = {3,3};
IppiPoint kernelAnchor = {1,1};

ippiFilter32f_8u_C1R( src+srcStep+1, srcStep, dst+dstStep+1, dstStep, dstSize, kernel, kernelSize,kernelAnchor);

Regards,
Vladimir

0 Compliments
szakatszozo
Débutant
773 Visites
ok, i will try this, but i would need to construct the kernel in way to use the info about the motion bluri know apriori: the number of pixels it moves and the angle;
in my case the number of pixels is the important one, the movement itself is always a vertical one...
could you help me in this?
thank you,
zozo'
0 Compliments
Répondre