滤波预处理在分水岭分割算法中的应用

Application of Filter Preprocessing in Watershed Segmentation Algorithm

影像中存在的无意义细节,导致在标记提取过程中会产生不必要的标记,加深分水岭分割的过分割问题。本文在标记提取前引入滤波预处理过程,消除噪声并平滑部分无意义细节。不同滤波由于其自身特性会对影像分割造成不同影响,分别使用中值滤波,均值滤波,双边滤波进行试验,研究滤波预处理对分割结果的影响,以及三种不同滤波的平滑效果对分割结果造成的影响的区别。并且进行了数值实验,证明了所提出方法的有效性。

The meaningless details in the image result in unnecessary markers in the process of marker extraction,which deepens the problem of over-segmentation in watershed segmentation.In this paper,the filter preprocessing process is introduced before the mark extraction,and the noise is eliminated and the part is meaningless.Different filtering due to its own characteristics will affect the image segmentation,respectively,using median filtering,mean filtering,bilateral filtering to study the effect of filtering preprocessing on the segmentation results,and three different filtering smooth effects on the segmentation results.Finally,numerical experiments are carried out to verify the effectiveness of the proposed method.