Image Ｒegistration Algorithm of SIFT Based on Edge and Corner Point

doi:10.3969 /j.issn.1674-0696.2013.04.40

Journal of Chongqing Jiaotong University(Natural Science) ›› 2013, Vol. 32 ›› Issue (4): 721-724.DOI: 10.3969 /j.issn.1674-0696.2013.04.40

Previous Articles Next Articles

Image Ｒegistration Algorithm of SIFT Based on Edge and Corner Point

Zhao Mengmeng，Cao Jianqiu

School of Information Science ＆ Engineering，Chongqing Jiaotong University，Chongqing 400074，China

Received:2012-05-28 Revised:2013-04-23 Online:2013-08-15 Published:2014-10-27

基于边缘角点的 SIFT 图像配准算法

赵萌萌，曹建秋

重庆交通大学信息科学与工程学院，重庆 400074

作者简介:赵萌萌(1989—),女,河南周口人,硕士研究生,主要从事图像处理、模式识别方面的研究。E-mail:583280590@qq.com。
基金资助:
重庆市科委攻关项目( CSTC 2011AC6102) ; 交通运输部西部项目( 20113188141480) ; 重庆交通大学创新基金项目〔2010( 下) 第16 号〕

Abstract

Abstract: SIFT is composed of feature extraction，feature descriptor representation and feature matching，which has a huge quantity in feature extraction and a high computational complexity in descriptor representation． So this algorithm is of low efficiency． In order to solve this problem，a kind of algorithm which is named SEC ( SIFT-Edge-Corner) is proposed． Corner points instead of SIFT feature points are extracted in image scale space． According to the theory that corner is the extreme curvature of edge，SEC constructs edge-of-Gaussian pyramid by canny operator before extracting corner points，and then the scales are selected． The experimental results show that SEC improves the feature extraction efficiency at the premise of ensuring high accuracy．

Key words: Scale Invariant Feature Transform ( SIFT) , edge extraction, corner point, SIFT-Edge-Corner ( SEC)

摘要： SIFT 由特征提取，特征描述符描述和特征匹配 3 部分构成，该算子特征提取数目庞大，建立特征描述符运算量高，导致算法效率低。提出了一种 SEC( SIFT-Edge-Corner) 算法，在图像尺度空间提取角点代替 SIFT 特征点，并根据角点是边缘曲率极值理论，预先采用 Canny 算子得到高斯边缘图像金字塔，再提取角点并进行尺度选择。实验结果表明: 该算法在保障高准确率的前提下大幅度提高特征提取效率。

关键词: SIFT, 边缘提取, 角点, SEC

CLC Number:

TP391.41

Zhao Mengmeng，Cao Jianqiu. Image Ｒegistration Algorithm of SIFT Based on Edge and Corner Point[J]. Journal of Chongqing Jiaotong University(Natural Science), 2013, 32(4): 721-724.

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.

References

[1] 裴聪，戴立玲，卢章平．基于sift 的简化算法下图像快速匹配[J]．制造业自动化， 2010，32( 1) : 132-135．
Pei Chong，Dai Liling，Lu Zhangping． A simplified algorithm based on sift for the quick image matching[J]． Manufacturing Automation，2010，32( 1) : 132-135．
[2] Ke Y，Sukthankar Ｒ． PCA-SIFT: A More Distinctive Ｒepresentation for Local Image Descriptors [C]/ / Proceedings of the 2004 IEEE Conference on Computer Vision and Pattern Ｒecognition．Washington D． C．，USA: IEEE CPS，2004: 511-517．
[3] Wang W X，Xu H L，Luo D J． Image Auto-Ｒegistration on Harris-Laplace Features [C]/ /3rd International Symposium on Intelligent Information Technology Application． Nanchang: IEEE CPS，2009:1121-1122．
[4] Lowe D G． Distinctive image features from scale-invariant key points [J]． International Journal of Computer Vision，2004，60( 2) : 91-110．
[5] 叶鹏，王琰．基于融合图像特征的角点提取方法[J]．沈阳理工大学学报，2010，29( 2) : 74-76．
Ye Peng，Wang Yan． Corner detection based on integrating image feature [J]． Journal of Shenyang Ligong University，2010，29( 2) : 74-76．
[6] 赵文彬，张艳宁．角点检测技术综述[J]．计算机应用研究，2006( 10) : 17-19．
Zhao Wenbin，Zhang Yanning． Survey on corner detection [J]．Application Ｒesearch of Computers，2006( 10) :17-19．
[7] Mokhtarian F，Mackworth A K． A theory of multi-scale，curvaturebased shape representation for planar curves [J]． IEEE Transaction on Pattern Analysis and Machine Intelligence，1992，14( 8) :789-805．
[8] He X C，Yung N H C． Curvature Scale Space Corner Detector with Adaptive Threshold and Dynamic Ｒegion of Support [C]/ / The 17th International Conference on Pattern Ｒecognition． Cambridge:IEEE CPS，2004: 791-794．
[9] Canny J A． Computational approach to edge detection [J]． IEEE Transaction on Pattern Analysis and Machine Intelligence，1986，8( 6) : 679-714．
[10] 曹建秋，王华清，蓝章礼．基于改进YCrCb 颜色空间的肤色分割[J]．重庆交通大学学报: 自然科学版，2010，29 ( 3) : 488-491．
Cao Jianqiu，Wang Huaqing，Lan Zhangli． Skin color division base on modified YCrCb color space [J]． Journal of Chongqing Jiaotong University: Natural Science，2010，29( 3) : 488-491．
[11] 樊志华，王春鸿，饶长辉，等．基于Harris 角点量与相位相关的亚像素级图像配准方法[J]．计算机应用研究，2011，28 ( 2) :788-791．
Fan Zhihua，Wang Chunhong，Ｒao Changhui， et al． High accuracy subpixel image registration based on phase-only correlation of Harris strength [J]． Application Ｒesearch of Computers，2011，28( 2) :788-791．

Image Ｒegistration Algorithm of SIFT Based on Edge and Corner Point

基于边缘角点的 SIFT 图像配准算法

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 6

Recommended Articles

Metrics

[1]	LI Shengyong1, ZHANG Zhihua1, WANG Shengnan2, WANG Meng2. Defect Recognition Algorithm for Vehicle Metal Materials [J]. Journal of Chongqing Jiaotong University(Natural Science), 2021, 40(07): 136-144.
[2]	SU Zhouyu, LAN Quanxiang, YUAN Quan, CAO Jianqiu. Potholes Image Edge Extraction Based on PCNN and Morphology [J]. Journal of Chongqing Jiaotong University(Natural Science), 2016, 35(1): 60-65.
[3]	Cao Jianqiu, Lan Quanxiang, Wang Danmei. Potholes Measurement Method Based on Image Processing Without Reference Material [J]. Journal of Chongqing Jiaotong University(Natural Science), 2015, 34(4): 57-61.
[4]	Feng Huanfei，He Youquan，Liu Chong. Adaptive Median Filter Based on Neighborhood Correlation [J]. Journal of Chongqing Jiaotong University(Natural Science), 2013, 32(3): 547-550.
[5]	DENG Tian-min, YU Yong, SHAO Yi-ming. A Novel Method of Vehicle Classification Based on Image Identification [J]. Journal of Chongqing Jiaotong University(Natural Science), 2008, 27(6): 1142-1145.
[6]	DU Ting-na. Morphological operation and structure element analysis about 2-numeric image [J]. Journal of Chongqing Jiaotong University(Natural Science), 2006, 25(1): 162-164.