The iterative detection network based recovery of black & white pictures shot by camera (implementation and complexity reduction issue)

The paper deals with a defocusing and Poisson (shot) noise elimination out of black & white pictures captured by a CCD (CMOS) camera with imperfectly adjusted lens. For purposes of this image recovery we can use the MAP criterion based iterative detection network (IDN) containing a number of mutually concatenated elementary function blocks so-called soft inversions (SISOs). This cellular structure makes IDN suboptimal but also numerically very simple and practically applicable in contrast to unviable optimal (single-stage) MAP detector. Firstly we focus closer on SISO entities and consequently on the creation of entire IDN. There will be introduced two different forms of reconstruction system (distributed IDN) together with complexity reduction possibilities and example of the image reconstruction by one of them.