A blind image restoration for out-of-focus blurred images using adaptive inverse filters
Tetsuya Imai, Ryoji Miyakoshi, Hironori Yamauchi · 2005
Abstract: This paper presents a new restoration approach to blurred images using inverse filters optimized by a Genetic Algorithm (GA). The proposed approach consists of two procedures. The first is to estimate a degradation level of blurred images based on Discrete Fourier Transform (DFT) and the next is to optimize inverse filters using a Real-coded GA. The degradation level estimation prior to inverse filters optimization enables the adaptive and efficient image restoration according to the blurred level. Experiments on natural images demonstrate that the proposed approach can provide a practical solution for a blind image restoration.