摘要： 针对常规的反锐化掩膜方法对图像进行增强时会放大噪声的问题，提出一种基于马尔科夫随机场理论的图像锐化方法，将Thin Plate先验能量函数引入到反锐化掩膜方法中。实验结果表明，该方法在突出图像边缘细节的同时，可有效抑制噪声。主观视觉效果和客观评价结果表明，该方法的处理效果较好。

关键词: 马尔科夫随机场, 图像锐化, 先验能量函数

Abstract: When normal unsharp mask method algorithm is applied to sharpening enhancement, the obvious noises are also enhanced. Based on this question, a method of image sharpening approach is proposed which is founded on the Thin Plate prior energy theory of Markov random field. The method can improve edge meanwhile it can’t enhance noise. The processing results of this method not only on the aspect of subjective visual effect but also on the aspect of objective evaluation are more satisfactory.

Key words: Markov random field, image sharpening, prior energy function

中图分类号: 

• TP391.41