Computer Engineering and Applications ›› 2011, Vol. 47 ›› Issue (16): 188-190.

• 图形、图像、模式识别 • Previous Articles     Next Articles

Smooth curve segment extracting method guided by gradient direction of boundary point

HU Hai’ou,ZHU Jianzhong   

  1. School of Information Science and Engineering,Hangzhou Normal University,Hangzhou 310036,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-06-01 Published:2011-06-01

一种边点梯度方向引导的光滑边段提取方法

胡海鸥,祝建中   

  1. 杭州师范大学 信息科学与工程学院,杭州 310036

Abstract: The extracting of boundary contour is the base of studying the shape characteristics.It has a major impact on the high level of image processing such as image characteristic description,identification and understanding.However,due to the effects of noise,light and many other reasons,the edge image obtained after pretreated often occurs the originally connected boundaries partial rupture.The traditional boundary tracing algorithm can not cross the breakpoints to continue tracing.In addition,when the branch occurs in the tracing process,the traditional boundary tracing algorithm,only to trace the first searched boundary point,can not control the right direction of boundary tracing.This paper presents a smooth curve segment extracting method guided by the gradient direction of boundary point,which not only can cross the breakpoint to continue tracing,but also can effectively reduce the wrong choice of the branch,while still meeting the requirement of a smooth edge segment.

Key words: boundary tracing, chain code, the gradient direction of boundary point, curve segment extracting, breakpoint, branch

摘要: 边界轮廓线的提取是研究形状特征的基础,它对图像特征描述、识别和理解等高层次的处理有重要作用。但由于噪声、光照等多种因素的影响,经过预处理得到的边缘图像往往会出现原本连接的边界局部断裂的现象,使传统的边界跟踪算法无法越过这些断点继续向下跟踪。另外,在跟踪过程中出现岔路时,传统的边界跟踪算法只跟踪第一个搜索到的边界点,无法控制边界的正确走向。提出了一种边点梯度方向引导的光滑边段提取方法,不仅能够跨越边界上的断点继续跟踪,而且能够有效地减少岔路选择错误的发生,同时还满足边段的光滑性要求。

关键词: 边界跟踪, 链码, 边点梯度方向, 边段提取, 断点, 岔路