基于云理论的图像分割新方法

针对区域生长算法的种子点初始化和生长准则问题,提出了一种基于云模型和区域生长的图像分割方法.该方法首先利用云变换对图像进行变换生成云模型,然后利用云模型的Ex作为区域生长的种子点,以云的极大判定法则作为区域生长准则进行区域生长,获得最终的分割结果.这种分割方法不但克服了区域生长法种子点和生长准则选取不当而产生过分割和欠分割的不足,而且很大地提高了云模型分割图片的速度,实验结果证明,该方法可以准确地分割出目标,是一种有效的图像分割方法.

A new method of image segmentation based on cloud theory

A segmentation algorithm based on the cloud model and the region growing theory is proposed. First, an image is transformed to a cloud model by cloud transformation, and then, the cloud grows regionally, taking Ex of the cloud model as the seed of region growing and enormous determination principle as the criteria of regional growing. In the end the final segmentation result can be got. This segmentation method can not only overcome excessive or less segmentation due to improper selection of seeds and the r...