综合区域和边界信息的图像自适应分割技术

Adaptive Image Segmentation Combining Region and Boundary Information

针对分水岭图像分割算法中的过分割问题,提出了一种综合区域和边界信息的解决技术.该技术主要利用区域和边界的综合信息,对分水岭算法处理后的过分割区域进行聚合.在基于区域和边界信息的聚合过程中,借鉴人眼视觉模型的韦伯感知原理,针对区域的不同亮度环境,自适应地选取动态的聚合阈值;并根据强弱边界属性调节聚合阈值,以鼓励对象内区域聚合和避免对象之间的区域聚合.试验结果表明,这种技术对分水岭算法中的过分割有较好的改善.

An efficient method was proposed to solve the over-segmentation problem of watershed segmentation algorithm, which is based on the region merger combining region and boundary information. In the merger processing, an adaptive threshold between the neighboring regions is used as the merging criteria, which is based on the Webber perception principle in human visual model, and the gradient is used to judge the true boundary of the image to avoid the merging regions of difference object. The experimental results show that this method can efficiently solve the over-segmentation problem.