Intel® Integrated Performance Primitives (Intel® IPP) Developer Guide and Reference
A newer version of this document is available. Customers should click here to go to the newest version.
Visible to Intel only — GUID: GUID-D63510FD-87CE-41E3-B2FE-AC6FD06D6430
Visible to Intel only — GUID: GUID-D63510FD-87CE-41E3-B2FE-AC6FD06D6430
Deconvolution
Functions described in this section perform image deconvolution. They can be used for restoring the degraded image, in particular image that was obtained by applying the convolution operation with known kernel. The Intel IPP functions implement two methods: the Fourier deconvolution (noniterative method) [see for example, [Puetter05]], and the Richardson-Lucy method (iterative method) [Ric72]. Border pixels of a source image are restored before deconvolution.
- DeconvFFTGetSize
Computes the size of the state structure for deconvolution with the fast Fourier transform (FFT). - DeconvFFTInit
Initializes the FFT deconvolution state structure. - DeconvFFT
Performs FFT deconvolution of an image. - DeconvLRGetSize
Computes the size of the state structure for Lucy-Richardson (LR) deconvolution. - DeconvLRInit
Initializes the LR deconvolution state structure. - DeconvLR
Performs LR deconvolution of an image.