• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2015, Vol. 37 ›› Issue (02): 384-389.

• 论文 • 上一篇    下一篇

一种图像的连续性边缘提取方法

贾迪1,董娜1,孟祥福1,李思慧1,陈硕2   

  1. (1.辽宁工程技术大学电子与信息工程学院,辽宁 葫芦岛 125105;
    2.国网辽宁省电力有限公司信息通信分公司,辽宁 沈阳 110006)
  • 收稿日期:2013-09-17 修回日期:2013-10-26 出版日期:2015-02-25 发布日期:2015-02-25
  • 基金资助:

    国家自然科学基金资助项目(61003162);辽宁省博士启动基金资助项目(20141142);辽宁工程技术大学生产技术问题基金资助项目(14T018);辽宁省大学生创新创业训练计划项目资助项目(201410147067);辽宁省教育2013年杰出青年学者成长计划资助项目(LJQ201303)

A method for continuous image edge extraction 

JIA Di1,DONG Na1,MENG  Xiangfu1,LI Sihui1,CHEN Shuo2   

  1. (1.School of Electronic and Information Engineering,Liaoning Technical University,Huludao 125105;
    2.Information and Communication Brach,Liaoning Electric Power Company Limited,Shenyang 110006,China)
  • Received:2013-09-17 Revised:2013-10-26 Online:2015-02-25 Published:2015-02-25

摘要:

图像边缘提取是图像处理领域的重要环节,边缘提取的连续性显得尤为重要,其结果可以作为语义分割、测量的基础,为此提出一种图像的连续性边缘提取方法。首先采用卡尼(Canny)算法完成图像的边缘检测,得到粗分割结果。由于该类算法采用算子进行操作,不可避免地产生非连续性区域,因此利用两类模板对粗分割结果进行形态学膨胀操作,连续性边缘采用小半径膨胀,边缘末端采用大半径膨胀,以此构成一幅膨胀图连接这些非连续性区域。最后,将该膨胀图作为平均曲率运动(MCM)方程嵌入函数u的初始值,通过引入g函数在原图中进行迭代求解,完成图像边缘的连续性提取。实验结果验证了算法的实用性与有效性。

关键词: 边缘提取, 连续性, Canny, 膨胀, MCM

Abstract:

Since edge extraction is an important step in image processing, continuity of edge extraction is particularly important.The extraction results can be applied to semantic segmentation and measurement.We present a method for continuous image edge extraction. Firstly,the Canny algorithm is used to get the coarse segmentation results.Because of the arithmetic operations,discontinuity region would be produced inevitably.Two classes of template expansion operations are used to get coarse segmentation results:small radius expansion should be selected for continuous region while bigger radius for the end of the edges,so an expansion graph is achieved. Finally,the expansion graph could be used as the initial u value for mean curvature motion (MCM) equation, the continuous edges will be extracted through iteration which introduces g function.The experimental results verify the practicability and effectiveness of the proposed algorithm.

Key words: edge extraction;continuity;Canny;expand;MCM