针对数学形态学骨架化算法可能出现的多像素宽度情况，定义了端点、节点、尾支和毛刺等概念，在此基础上设计了一种分级毛刺去除算法。
Aimed at the appearance of multi-pixels width of skeleton, this paper gave the definition of tip, node, tail and burr, and a design of an algorithm for iteration burr.
由于在线去除焊管内毛刺的方法造成焊管废品率较高，质量难以保证。
As online within the burr removal method pipe caused Welded pipe reject rate higher, it is difficult to guarantee quality.
提出一种“高斯滤波”的方法去除这些白点，使二值化图像平滑、少毛刺，并具有较好的连通性。
But after the Gauss filter process, the threshold image is changed into smooth one which has smaller burr and has better connectedness.
应用推荐