摘要
提出了一种新的二值图象复原算法，算法基于Ｍａｒｋｏｖ场模型。算法在象素值的更新上采用了寻找极大似然解的方法，同时引入了概率接受机制，通过计算接受函数来判定是否接受新值。算法具有很高的计算效率，在复原中取得了良好的效果。
In this paper, a new binary image restoration algorithm is proposed, which is based on Markov Random Fields. This algorithm finds the ML solution in the procedure of pixel value updating. The mechanism of probability receptence is introduced to judge the transition of pixel value by computing the receptence function. Its computation efficiency is very high. Good results are gained in the experiments.
出处
《中国图象图形学报（A辑）》
CSCD
1999年第7期557-561,共5页
Journal of Image and Graphics
基金
国家自然科学基金
