基于图论阈值算法的图像分割研究

Image mosaic research based on wavelet and rough set algorithm

为了提高阈值分割图像的质量,提出了采用图论阈值算法。首先,构造图论和图像的映射函数关系,每个顶点通过点来映射,每条边通过线来映射。用基于区域属性的图像边缘决策表,不同像素点或不同组像素点之间的灰度特征差作为权重系数,通过基于决策属性权重来构造像素联系图;然后,采用聚类法计算像素到目标类和背景类的相似程度,最小生成树策略解决伪割集问题;最后,给出图像阈值设定以及算法流程。实验仿真表明,本文算法的分割图像效果清晰,消除了图像分割中存在的过合并和欠合并现象,本文算法的信息熵为28.780 3bit,处理时间为1.454 3s。满足分割结果中对执行时间少、信息含量大等要求。

In order to improve the quality of image segmentation,Graph Theory Threshold Algorithm is established.First,map function is constructed between the graph and image,each vertex is mapped by point,each edge is mapped by line.Second,decision table of image edge is set by attribute of area,gray feature is defined weight coefficient between pixel of different points and groups,decisions of attribute weight are constructed with figure of pixel.Third,similar of pixel is computed by cluster between object and background class,minimum spanning tree method solved the problem of pseudo cut set.Finally,process is described.The experiments show this algorithm made image edge clearly,eliminate the over and under of merger phenomenon,entropy of information is 28.7803 bit,time are 1.454 3 s.It takes less time and has large information.