计算机图象分割及其算法

The Image Partition and Algorithm of the Computer

从图象应用角度出发,利用象点的相似性和非连续性原理,对灰度取阈值进行图象分割.选择阈值采用两种方法一种是利用图象统计量——矩守恒原理进行选择,并且给出二值化法和多值化法的闽值选择过程及算法;另一种是利用分类图象的灰度等级方法进行选择,并且给出分类法的计算机算法.

From the applicational points of view,image partition can be done by selecting a threshold of grey levels using the similarity and uncontinuity of pixeIs.In case of simple images,we use the binary method to select a thre- shold to partition an image.In case of a complicated image,we use the multiple method to select several thresholds to partition the image.There are two methods in threshold partition.One is the partition on the law of conservation of quantum-moment,and the threshold selecting procedures and algorithms for binary method and multiple threshold selecting method are pre- sented;the other method is to use grey levels of various images to select the threshold,and the algorithms for image categoriztion is also provided.