基于块遍历的直线边缘特征提取

张建伟; 张启衡

doi:null

您当前的位置：

首页 >

文章列表页 >

基于块遍历的直线边缘特征提取

信息科学 | 更新时间：2020-08-12

- 基于块遍历的直线边缘特征提取
- Line-edge feature extraction based on ergodic block
- 光学精密工程 2009年17卷第3期页码：662-668
- 作者机构：
  
  1. 中国科学院研究生院北京,100039
  2. 中国科学院光电技术研究所国家863计划光束控制重点实验室,四川成都,610209
- 作者简介：
- 基金信息：
  
  国家863高技术研究发展计划资助项目(No.2007AA802401)()
- DOI：
  中图分类号： TP391.4
- 收稿日期：2008-05-22，
  
  修回日期：2008-07-24，
  
  网络出版日期：2009-03-25，
  
  纸质出版日期：2009-03-25
- 稿件说明：
移动端阅览
张建伟, 张启衡. 基于块遍历的直线边缘特征提取[J]. 光学精密工程, 2009,17(3): 662-668

ZHANG Jian-wei, ZHANG Qi-heng. Line-edge feature extraction based on ergodic block[J]. Editorial Office of Optics and Precision Engineering, 2009,17(3): 662-668
张建伟, 张启衡. 基于块遍历的直线边缘特征提取[J]. 光学精密工程, 2009,17(3): 662-668 DOI：

ZHANG Jian-wei, ZHANG Qi-heng. Line-edge feature extraction based on ergodic block[J]. Editorial Office of Optics and Precision Engineering, 2009,17(3): 662-668 DOI：

摘要

为了能够快速而精确地提取图像中扩展目标的直线边缘特征

提出了一种基于块遍历的直线提取方法。该算法先通过块遍历整个图像

然后按照直方图把块分类并进行二值化

再通过链码得到相同单像素直线的多个块表示

然后通过这些直线的斜率和截距进行分类

最后对分类结果求均值。仿真结果表明

该方法不仅能正确地提取直线特征且提取的直线可以达到亚像素精度

其角度定位精度超过0.01 rad。从运算量和复杂性来看

其综合性能要优于经典霍夫变换方法

可以得到很好的直线提取结果

并为硬件实现打好基础。

Abstract

In order to extract the line-features of extended object more quickly and more accurately

a line-edge extraction method is proposed based ergodic block. With proposed algorithm

a whole image is traversed by blocks

then the blocks with double peaks is obtained according to the property of gray-histogram.After processing those blocks by gray threshold and chain-edge extraction

the blocks without line characters are eliminated

the others with line characters are sorted by slope and intercept.Finally

the lines are got by computing the mean of similar slopes and intercepts. Experiment results show that proposed algorithm can extract line in subpixel precision and the precision of angle is higher than 0.01 rad. The complexity and computation of proposed algorithm is more better than that of Hough transform

which provides a way for hardware-implementation.

关键词

Keywords

references

付树军,阮秋琦,穆成坡,等. 基于非线性扩散滤波的边缘检测和图像测量[J]. 光学精密工程, 2007,15(2):289-293. FU SH J, RUAN Q Q, MU CH P, et al.. Edge detection for image measurement based on nonlinear diffusion filtering[J]. Opt. Precision Eng., 2007,15(2): 289-293. (in Chinese)[2] ILLINGWORTH J ,KITTLER J. A survey of the Hough transform [J]. Computer Vision, Graphics and Image Processing, 1988, 44:87-116.[3] AGGARWAL N , KARL W C. Line detection in images through regularized Hough transform [J].IEEE Image Processing, 2006, 15(3):873-876.[4] PERANTONIS S J, VASSILAS N, TSENOGLOU T, et al.. Robust line detection using weighted region based Hough transform [J]. Electronics letters, 1998, 34(7):648-650.[5] 鲁光泉,许洪国,李一兵. 基于链码检测的直线段检测方法[J]. 计算机工程, 2006,32(14):1-3,10. LU G Q, XU H G, LI Y B. Line segment detection based on chain code detection . Computer Engineering, 2006 ,32(14):1-3,10. (in Chinese)[6] 于新瑞,王石刚,王高中,等. 数字图像直线特征的亚像素位置检测[J]. 光学技术,2004,30(2):138-141. YU X R, WANG SH G, WANG G ZH, et al.. Subpixel location detection of line feature in digital image [J]. Optical Technique, 2004,30(2):138-141. (in Chinese)[7] 何博侠, 张志胜, 戴敏,等. 基于序列局部图像的高精度测量[J]. 光学精密工程,2008,16(2):367-373. HE B X, ZHANG ZH SH, DAI M, et al.. A high-precision dimension measurement method based on sequential partial images [J]. Opt. Precision Eng., 2008,16(2):367-373. (in Chinese)

浏览量

354

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于分数阶微分的高光谱图像特征提取与分类

面向高光谱影像场景分类的轻量化深度全局-局部知识蒸馏网络

融入一维概率Hough变换与局部Zernike矩的双目视觉测量

多分支无锚框网络密集行人检测算法

SENet生成对抗网络在图像语义描述中的应用