一种基于快速鲁棒特征的图像匹配算法
收稿日期: 2010-04-28
修回日期: 2010-06-28
网络出版日期: 2011-02-25
基金资助
2009年西北工业大学研究生创新基金（Z200937）及敦煌研究院“数字敦煌”项目的支持
An Improved Algorithm of Image Registration Based on Fast Robust Features
Received date: 2010-04-28
Revised date: 2010-06-28
Online published: 2011-02-25
针对传统的图像特征匹配算法数据量大、计算耗时长的缺点，本文提出了一种基于快速鲁棒特征(SURF)的图像配准算法。SURF算法作为一种新的特征提取算法，在独特性、鲁棒性等方面均超过了其它方法，并在计算效率上具有明显的优势。该算法在积分图像的基础上进行快速计算，通过快速Hessian检测子来检测特征点。对于每个特征点，通过计算哈尔小波变换来确定特征点的主方向，并确定特征描述子，再根据Hessian矩阵迹的正负性和最近邻与次近邻比值的方法相结合获取匹配点，并用改进的RANSAC算法剔除伪匹配点以确保匹配的有效性。实验表明，该算法既能满足匹配准确性的要求，又具有计算量小、计算速度快的优点。
王君本，卢选民，贺兆 . 一种基于快速鲁棒特征的图像匹配算法[J]. 计算机工程与科学, 2011 , 33(2) : 112 -117 . DOI: 10.3969/j.issn.1007130X.2011.
With the shortcomings of large data amount and long time consuming in the conventional image feature matching algorithms, a new algorithm based on SURF for image registration is presented in this paper. SURF is a new feature extraction algorithm which approximates or even outperforms other schemes with respect to distinctiveness, and robustness. And it is much faster than the other similar ones. It is fast computed based on the integral image and through the FastHessian detector, the feature points are extracted. For each feature point, the dominant orientation is assigned by computing the Haarwavelet responses, and then the descriptor is generated. Image matching is made based on the sign of the trace of the Hessian matrix and the ratios of the closest neighbor and the second closest neighbor, and an improved RANSAC technique is applied to eliminate outliers to ensure the effectiveness of the matched pairs. The experimental result shows that this algorithm can not only meet the requirement of accuracy, but also has a small data amount and fast speed for image registration.
［1］Harris C, Stephens M J.A Combined Corner and Edge Detector［C］∥Prco of the 4th Alvey Vision Conf, 1988:147152.
［2］Lowe D G. Distinctive Image Features from ScaleInvariant Keypoints［J］. International Journal of Computer Vision, 2004, 60(2):91110.
［3］Brown M,Lowe D G. Invariant Features from Interest Point Groups［C］∥Proc of British Machine Vision Conf, 2002:656665.
［4］Savb J, Krajnik T, Faigl J,et al.FPGA Based Speeded Up Robust Features［C］∥Proc of the IEEE Int’l Conf on Digital Object Identifier,2009:3541.
［5］Bay H, Tuyteplaars T, van Gool L. SURF:Speeded Up Robust Features［C］∥Proc of European Conf on Computer Version, 2006:404417.
［6］Bay H, Ess A,Gool L Van. SpeedUp Robust Features(SURF)［J］. Computer Vision and Image Understanding,2008,110(3):346359.
［7］Valgren C, Lilienthal A. SIFT SURF and Seasons: Longterm Outdoor Localization Using Local Features［C］∥Proc of the 3rd European Conf on Mobile Robots, Freiburg, Germany, 2007:253260.
［8］Viola P, Jones M. Rapid Object Detection Using a Boosted Cascade of Simple Feature［C］∥Proc of CVPR’01,2001:511518.
［9］Lindeberg T. Feature Detection with Automatic Scale Selection［J］. IJCV, 1998，30(2):79116..
［10］Mikolajczyk K,Schmid C. A Performance Evaluation of Local Descriptors［J］. IEEE Trans on Pattern Anal Mach Intell,2005,27(10):16151630.
［11］邵平， 杨路明，曾耀荣. 计算旋转Harr型特征的积分图像的算法改进［J］. 计算机技术与发展,2006,16(11):146147.
［12］朱松立， 戴礼荣,宋彦,等.基于角点特征值和视差梯度约束的角点匹配［J］. 计算机工程与应用， 2005,41(34):6264.
［13］Liu Ruihua,Wang Yanguang. SAR Image Matching Based on Speeded Up Robust Feature［C］∥Proc of WRI Global Congress on Intelligent Systems,2009:518522.
/
| 〈 |
|
〉 |