摘要
针对图像分割边缘不准确的问题,研究了一种基于模糊理论的阈值区间的图像分割方法。在首先介绍的模糊阈值分割的基本原理上,提出了一种分层分割图像的思想。根据图像具有模糊的性质,利用模糊阈值法得到一个图像分割的调和阈值,再将每一层根据像素统计直方图信息得到一个本层次的阈值区域,最后用模糊阈值法得到的阈值调和阈值区域,使最终的分割阈值区间更精确。最后,根据相邻层相连背景像素相似的原则,逐层分割图像。实验结果表明该方法具有较好的分割效果。
Aimed at the problem that image segmentation is not accurate,research a method of image segmentation is based on fuzzy of the thresholding region.At first introduce the basic principles of fuzzy thresholding on segmentation.A hierarchical image segmentation idea is presented.According to the fuzzy nature of images obtain an attemper thresholding of image segmentation.Histogram to expose the pixels distributing in every layer,with the information of pixel statistic to figure out every layer＇s background pixel region.Finally,using the fuzzy thresholding attemper the thresholding region.By the relationship between adjoin layers of similar background pixels,segment layers.The experimental results demonstrate this kind of algorithm is effective and preferable.
出处
《计算机技术与发展》
2010年第5期121-123,127,共4页
Computer Technology and Development
基金
江苏省计算机信息处理技术重点实验室开放课题项目(08KJB520003)
关键词
像素统计
图像分割
阈值区域
调和阈值
自适应
pixel statistic
image segmentation
threshold region
attemper threshold
adaptive
