一种基于DA-GMRF的无监督图像分割方法

Unsupervised image segmentation based on DA-GMRF

提出一种基于间断自适应高斯马尔可夫随机场(DA-GMRF)模型的无监督图像分割方法。针对MRF模型中的过平滑问题,利用边缘信息构造能量函数,定义了一种DA-GMRF模型。利用灰度直方图势函数自动确定分类数及分割阈值,进行多阈值分割,得到DA-GMRF模型中标记场的初始化,用Metroplis采样器算法进行标记场的优化,实现了图像的无监督分割。实验结果表明了该方法的有效性。

An unsupervised image segmentation method basedon DA-GMRF （Discontinuity- Adaptive Gaussian Markov Random Field） model is proposed. To solve the drawback of over-smoothness in the MRF model, a kind of DA-GMRF model is defined, in which the edge information of image is used to construct corresponding energy functions. In order to start the unsupervised image segmentation, a multi-threshold image segmentation algorithm is presented based on potential function of gray histogram to initialize the label field. This algorithm can determine region number and multi-thresholds automatically. At last, Metroplis sampler algorithm is adopted to optimize the label field. Segmentation experiments on several images show that the algorithm proposed is effective.