PDF Estimation in the Blind Deconvolution Algorithm Based on Minimum Output Entropy
Ju Liu · Journal of Shandong University · 2003
Two pdf estimation methods are given in this paper. The pdf is used in minimum entropy based multi channel blind deconvolution algorithm. One of the estimation methods is based on the Edgeworth expansion and the other is based on the parameter approximation of the kernel function methods. The blind deconvolution algorithm is consisted of the pdf estimation and the learning rule of the filtering parameter in deconvolution. Computer simulation shows the separation ability of the algorithm.