|本期目录/Table of Contents|

[1]甄巍松,李国强,鲁统伟.基于特征点相似度的匹配定位算法[J].武汉工程大学学报,2011,(04):85-88.[doi:10.3969/j.issn.16742869.2011.04.022]
　ZHEN Wei song,LI Guo qiang,LU Tong wei.Match and location algorithm based on similarity of feature point[J].Journal of Wuhan Institute of Technology,2011,(04):85-88.[doi:10.3969/j.issn.16742869.2011.04.022]
点击复制

基于特征点相似度的匹配定位算法(/HTML)
分享到：

《武汉工程大学学报》[ISSN:1674-2869/CN:42-1779/TQ]

卷:
期数:
2011年04期
页码:
85-88
栏目:
机电与信息工程
出版日期:
2011-04-30

文章信息/Info

Title:
Match and location algorithm based on similarity of feature point
文章编号:
16742869(2011)04008504
作者:
甄巍松1; 李国强1; 鲁统伟2
1.光电控制技术重点实验室,河南 洛阳 471009;
2.武汉工程大学智能机器人湖北省重点实验室，湖北 武汉 430074
Author(s):
ZHEN Weisong1;  LI Guoqiang1;  LU Tongwei2
1.Science and Technology on Electrooptic Control Laboratory, Luoyang 471009, China;
2.Hubei Province Key Laboratory of Intelligent Robot, Wuhan Intitute of Technology, Wuhan 430074, China
关键词:
尺度不变; 极线约束; 匹配; 定位
Keywords:
scaleinvariant;  epipolar constraint;  matching;  location
分类号:
TP391
DOI:
10.3969/j.issn.16742869.2011.04.022
文献标志码:
A
摘要:
前视定位系统中可能存在如视点、方位和距离等误差,导致在匹配时刻获取的实时目标场景与模板中目标不一致,从而影响目标定位的准确性.本文提出了一种基于特征点相似度的匹配定位算法,首先在图像的尺度空间上提取尺度不变特征点.然后根据描述子来进行相似度的判别,得到初始的匹配点集,然后利用极线约束,从而消除匹配错误的点.利用类似RANSAC方法估计场景中目标的变换参数,从而确定场景中目标所在的位置、尺度变化和旋转角度.实验结果验证了该算法的有效性和鲁棒性.
Abstract:
The targets in the template and the realtime scene are different duo to the errors in the forwardlooking locating system such as the view point, orientation, distant et al, which affect the accuracy of the target location. This paper presents a matching and locating algorithm based on the similarity of the feature points. Firstly, the scaleinvariant feature points are extracted in the image scale space. Then the set of the initial matching points is build through the similarity of the descriptor. The wrong matching points are eliminated using the epipolar constraint. Finally, the position, scale factor and rotation angle are determined by estimating the parameters using RANSAC. Experimental results show the validity and robustness of the algorithm.

参考文献/References:

［1］鲁统伟.前视目标图像匹配定位技术研究［D］.武汉：华中科技大学图像所,2009.
［2］李相育，刘辉.图像处理技术在染色体自动分析中的应用［J］.武汉化工学院学报，2003，25（1）：7779.
［3］杨述斌，陈艳菲，程莉.交替滤波的加权形态边缘检测算法［J］.武汉工程大学学报，2009，31（7）：8890.
［4］刘兴，刘庆祥.一种实用的车牌图象获取方案［J］.武汉化工学院学报，2005，27（5）：8285.
［5］Harris C, Stephens M.A combined Edge and Corner Detector［J］.In 4th Alvey Vision Conference,1988:189192.
［6］Zhang Z, Deriche R, Faugeras O, et al.A robust technique for matching two uncalibrated images through the recovery of the unknown epipolar geometry［J］. Artificial Intelligence,1995,78:87119.
［7］Schmid C, Mohr R.Comparing and evaluating interest points［J］.In Proceedings of the International Conference on Computer Vision,1998:230235.
［8］Matas J, Koubaroulis D, Kittler J.Colour image retrieval and object recognition using the multimodal neighborhood signature［J］.In Proceedings of the European Conference on Computer Vision,2000:4864.
［9］Tuytelaars T, VanGool L. Wide baseline stereo based on local, affinely invariant regions［J］.In Proceedings of the British Machine Vision Conference,2000:121130.
［10］Baumberg. Reliable feature matching across widely separated views［J］.In Conference on Computer Vision and Pattern Recognition,2000:774781.
［11］Lowe D G. Distinctive image features from scaleinvariant keypoints［J］.International Journal of Computer Vision,2004,60(2):91110.
［12］Mikolajczyk K, Schmid C. A performance evaluation of local descriptors［J］.In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,2003(2):257263.
［13］陈泽志,吴成柯.一种高精度估计的基础矩阵的线性算法［J］.软件学报,2002,13（4）：840845.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期：20101122作者简介：甄巍松（1980），男，河南洛阳人，工程师.研究方向：嵌入系统设计、机器视觉.
更新日期/Last Update: