Wavelet Domain Image Restoration Algorithm using Classified Hidden Markov Tree Model
Yaping Zhu, Tingzhi Shen, Yu Cui · 2006
An image restoration algorithm based on classified hidden Markov tree (CHMT) model in the wavelet domain is proposed, faced with the non-stationary property of real-world images and the difficulty of computation problem in image recovery. In light of Bayesian framework of image restoration method, CHMT model is used as a priori information of image in the wavelet domain, and regularization restriction is made to recover image. The CHMT model has spatial adaptability, which makes the modeling more accurate. A classified simplified method is presented, to makes it easy to find solutions of recovery equation. Experimental results show that, this algorithm has a reasonable computational complexity, can effectively preserve edge and pointed information, and the restoration quality is improved significantly