摘要
提出了一种基于直方图极大值准则图像阈值分割方法,通过调节参数可对多类图像进行分割。阈值分割是实时图像处理首选的分割算法。本文通过一维灰度直方图确定极大值,从而获得图像中多个目标区域和背景区域阈值,分割出多个区域,将每个区域赋为不同的灰度级,根据具体所要分割的目标再进行分割。实验证明该方法分割效果较好,分割速度快,实时性较好。
Based on histogram maximum criterion threshold image segmentation method is presented in this paper. Multiple thresholds of objects and background can be obtained from the 1-D gray level histogram. The image can be segmented based on the representative thresholds. After image segmentation, some regions can be segmented, and different regions will be endowed with different gray values. The final object can be segmented from the image that only includes a few gray values. An excellent result can be obtained when the proposed method is used in the threshold image segmentation.
出处
《现代电子技术》
2005年第18期47-49,共3页
Modern Electronics Technique
关键词
图像分割
阈值
直方图
极大值
可变参数
image segmentation
threshold
histogram
maximum criterion
alterable parameter
