图像分割的分层处理方案

Image segmentation based on multi-level technique

提出了一种分层处理的图像分割方案,该方案将分割过程分为两步:第一步,在低分辨率的条件下运用二维otsu阈值化方法对图像进行粗分割;第二步,在精细分辨率的条件下采用模糊C-均值算法对图像进行细分割,同时将第一步所得的结果融入第二步,以避免细分割中为求聚类中心所做的大量计算。实验结果表明,这一方案有利于提高分割质量和缩短分割时间。

A new multi -level based on image segm entation technique that is divided i nto two steps is presented.At first,the image is sengmented in a low level by using two -dimensional o stu thresholding to find the coarse contour of regions;then the fuzzy -C-means algorithm is used to process t he image in a high level to find its delicate character.To avoid the large amount of calculation,the result of the first step is fused into the sec ond step.The experimental results show that this technique is effective in i mproving the quality of segmentatio n and lessening the calculating time.