Multichannel image restorations using an iterative algorithm in space domain

摘要

The critical issue in extending the iterative algorithm widely used for single-channel restoration to the multichannel case is the formidable computational requirements due to the huge matrix operations in the restoration algorithm. This paper explores the possibility of combining the multichannel constrained-least-square (CLS) algorithm with the iterative algorithm for restoring multichannel images. A detailed analysis of the properties of the matrices in the formula and an effective method implemented in space domain are presented. This makes the iterative CLS algorithm feasible for multichannel restoration problems. Experiments are performed to evaluate the proposed iterative multichannel CLS algorithm. The results of the related independent single-channel algorithm are also presented for comparison.