基于模糊距离变换的骨架剪枝算法

针对传统骨架提取算法结果无法保证骨架单像素性,并且伴有毛刺产生的现象,提出了一种能够有效去除骨架中毛刺分枝,充分体现物体形态特征的骨架剪枝算法。该算法以模糊距离变换为理论基础,在物体粗骨架图像中计算每条尾枝的模糊距离变换值,使用根据骨架特点确定的动态阈值判断骨架分枝是否被去除,并利用分级剪枝方法实现骨架剪枝操作。实验结果表明,该算法在进行剪枝操作中可有效地保证骨架的准确性和连续性。

Skeleton pruning algorithm based on fuzzy distance transform

To solve the problem of multi-pixels and burr in traditional skeleton extraction algorithm,this paper presents a skeleton pruning algorithm which can effectively remove the burr and fully reflect the morphological characteristics of the object.The fuzzy distance transform based algorithm computes the fuzzy distance transform values of each tail in the crude skeleton,judges whether the skeleton branches should be removed according to the dynamic threshold determined by the characteristics of the skeleton and achieves the skeleton pruning by using the classification pruning method.The experiment results show that the algorithm can ensure accuracy and continuity of the skeleton during the pruning operation.