利用边缘直线特征的异源图像配准

方兰兰; 刘贵喜; 吕孟娇; 姚李阳; 张佳明

doi:10.3788/OPE.20152313.0753

您当前的位置：

首页 >

文章列表页 >

利用边缘直线特征的异源图像配准

更新时间：2020-08-13

- 利用边缘直线特征的异源图像配准
- Multi-source matching of multi-sensor images based on edge line features
- 光学精密工程 2015年23卷第10z期页码：752-760
- 作者机构：
  
  西安电子科技大学机电工程学院, 陕西西安 710071
- 作者简介：
- 基金信息：
  
  国防预研基金资助项目(No.Y420150401XX)();国家部委十二五科技规划资助项目(No.Y31011040315)();中央高校基本科研业务费专项资金资助项目
- DOI：10.3788/OPE.20152313.0753
  中图分类号： TP391.4
- 收稿日期：2015-06-02，
  
  修回日期：2015-06-20，
  
  纸质出版日期：2015-11-14
- 稿件说明：
移动端阅览
方兰兰, 刘贵喜, 吕孟娇等. 利用边缘直线特征的异源图像配准[J]. 光学精密工程, 2015,23(10z): 752-760

FANG Lan-lan, LIU Gui-xi, L&#220 etc. Multi-source matching of multi-sensor images based on edge line features[J]. Editorial Office of Optics and Precision Engineering, 2015,23(10z): 752-760
方兰兰, 刘贵喜, 吕孟娇等. 利用边缘直线特征的异源图像配准[J]. 光学精密工程, 2015,23(10z): 752-760 DOI： 10.3788/OPE.20152313.0753.

FANG Lan-lan, LIU Gui-xi, L&#220 etc. Multi-source matching of multi-sensor images based on edge line features[J]. Editorial Office of Optics and Precision Engineering, 2015,23(10z): 752-760 DOI： 10.3788/OPE.20152313.0753.

摘要

针对异源异构图像中无法利用特征点进行图像匹配的问题

提出了一种基于直线的图像匹配方法。首先

检测出图像的边缘点并连接成曲线

利用多尺度策略提取边缘直线段;然后

根据空间接近性和相对显著性

将边缘直线段进行分组

每个组都被看作是直线的一个特征

称其为直线标签;最后

根据线段的几何形状计算线段之间的相似性

从而判断两个直线标签之间是否相似

并利用快速匹配算法进行匹配。类似于局部特征

直线特征具有鲁棒性

适用于视角的变化。与现有的局部特征匹配方法相比

本文的直线方法更适用于缺乏纹理

纹理不清晰

光照变化很大以及异源异构的场景图像。实验结果表明

本文的算法具有尺度不变性、旋转不变性以及光照不变性

能有效地解决无法用点特征匹配算法的场景图像之间的匹配

匹配精度在1 pixel以内。

Abstract

A novel multi-source image matching strategy based on line feature extraction algorithm was proposed to overcome the shortcoming of point feature matching algorithm. Firstly

image edge points were detected and connected into curves

and the edge line segment was extracted by utilizing multi-scale strategy. Then

based on the spatial proximity and relative significance

the edge line segment was divided into lots of groups

and each of them was regarded as a feature of line called line labels. Finally

the similarity between the lines was computed according to the line geometry. Consequently

whether the two line labels are similar could be verified and then they were matched by the fast matching algorithm. Similar to local feature

the line feature is robust and adapted to changes of angle of view. Compared to matching approaches based on existing local features

the proposed method is more suitable for the images with low-texture scenes

blurring-texture scenes

high changing of illumination or multi-source scenes. Experimental results indicate that the proposed algorithm has scale invariant

rotate invariant and illumination invariant and can solve the situation in which point feature matching algorithm is not available and the matching accuracy is within one pixel.

关键词

Keywords

references

曾峦,王元钦,谭久彬. 改进的SIFT特征提取和匹配算法[J]. 光学精密工程,2011,19(6):1391-1397. ZENG L, WANG Y Q, TAN J B. Improved algorithm for SIFT feature extraction and matching[J]. Opt. Precision Eng., 2011, 19(6):1391-1397.(in Chinese)

翟优,曾峦,熊伟. 不同局部邻域划分加速鲁棒特征描述符的性能分析[J]. 光学精密工程,2013, 21(9):2395-2404. ZHAI Y, ZENG L, XIONG W. Performance analysis of SURF descriptor with different local region partitions[J]. Opt. Precision Eng., 2013, 21(9):2395-2404.(in Chinese)

巨西诺, 郭文普, 孙继银,等. 基于兴趣点的遥感影像可匹配性度量[J]. 光学精密工程, 2014, 22(4):1071-1077. JU X N, GUO W P, SUN J Y, et al.. Matching probability metric for remote sensing image based on interest points[J]. Opt. Precision Eng., 2014, 22(4):1071-1077.(in Chinese)

WANG Y, MA Y, CHEN Q. A method of line matching based on feature points[J]. Journal of Software, 2012, 7(7):1539-1545.

ZHU Y H, LI J SH, ZHU Y J. A matching algorithm for IR/visual images based on the invariant moment of histogram[C]. Seventh Symposium on Multispectral Image Processing and Pattern Recognition(MIPPR2011)-Pattern Recognition and Computer Vision, 2011:80040Q-80040Q-5.

谭东杰, 张安. 方向相关与互信息加权组合多模图像配准方法[J]. 红外与激光工程, 2013,42(3):836-841. TAN D J, ZHANG AN. Multi-model image registration based on weighted orientation correlation and mutual information[J]. Infrared and Laser Engineering, 2013,42(3):836-841.(in Chinese)

WANG B, LI Y, LU Q, et al.. Image registration algorithm based on modified GLOH descriptor for infrared images and electro-optical images[J]. Lecture Notes in Electrical Engineering, 2012:365-370.

BACLAWSKI A. Experimental relative line strengths within selected multiplets of neutral sulfur from the visible and infrared spectral range[J]. European Physical Journal D, 2011, 61(2):327-334.

IWAI D, SATO K. Optical superimposition of infrared thermography through video projection[J].Infrared Physics & Technology, 2010, 53(3):162-172.

VERHAGEN B, TIMOFTE R, van GOOL L. Scale-invariant line descriptors for wide baseline matching[C].2014 IEEE Winter Conference on Applications of Computer Vision(WACV), IEEE, 2014:493-500.

BAY H, FERRARI V, van GOOL L. Wide-baseline stereo matching with line segments[C]. 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2005:329-336.

WANG Z, WU F, HH Z. MSLD:A robust descriptor for line matching[J]. Pattern Recognition, 2009, 42(5):941-953.

赵小川, 罗庆生, 陈少波. 改进型图像中的直线快速检测[J]. 光学精密工程, 2010, 18(7):1654-1660. ZHAO X CH, LUO Q SH, CHEN SH B, Improved fast line detection method in images[J]. Opt. Precision Eng., 2010, 18(7):1654-1660.(in Chinese)

WANG L, YOU S Y, NEUMANN U. Supporting range and segment-based hysteresis thresholding in edge detection[J]. 2008 15th International Conference on Image Processing, 2008,1-5:609-612.

MELEN T, OZANIAN T. A fast algorithm for dominant point detection on chain-coded contours[J]. Lecture Notes in Computer Science, 1993:245-253.

赵萌萌, 曹建秋. 基于边缘角点的SIFT图像配准算法[J]. 重庆交通大学学报:自然科学版, 2013, 32(4):721-724. ZHAO M M, CAO J Q. Image registration algorithm of SIFT based on edge and corner point[J]. Journal of Chongqing Jiaotong University(Natural Science), 2013, 32(4):721-724.(in Chinese)

FERRARI V, FEVRIER L, JURIE F, et al.. Groups of adjacent contour segments for object detection[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2008, 30(1):36-51.

浏览量

263

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

采用多站图像直线特征的飞机姿态估计