基于GPU的SIFT特征匹配算法并行处理研究

计算机科学 ›› 2013, Vol. 40 ›› Issue (12): 295-297.

基于GPU的SIFT特征匹配算法并行处理研究

姜超,耿则勋,娄博,魏小峰,沈忱

解放军信息工程大学郑州450052;解放军信息工程大学郑州450052;解放军信息工程大学郑州450052;解放军信息工程大学郑州450052;解放军信息工程大学郑州450052

出版日期:2018-11-16 发布日期:2018-11-16
基金资助:
本文受863项目(2012AA7032031D)资助

Parallel Processing Research on SIFT Feature Matching Algorithm Based on GPU

JIANG Chao,GENG Ze-xun,LOU Bo,WEI Xiao-feng and SHEN Chen

Online:2018-11-16 Published:2018-11-16

摘要/Abstract

摘要： SIFT算法因具有旋转、缩放以及平移不变性而在影像配准和基于影像的三维重建领域得到广泛应用。但该算法复杂度较高,在CPU上执行的效率不高,难以满足对实时性要求较高的应用。在深入分析SIFT算法原理的基础上,针对该算法提取特征的多量性和特征向量的高维性,将该算法进行了并行化改造以利用GPU强大的并行计算能力,并与CPU上实现的SIFT算法进行了比较。实验证明,基于GPU的SIFT算法执行效率大幅提升,平均可以达到10倍以上的加速比。

关键词: GPU,SIFT,CUDA,特征匹配

Abstract: SIFT algorithm has invariance of rotation,scale and translation,so it is used widely in the field of image matching and 3D reconstruction．But the SIFT algorithm is complicate,making the processing speed slow,difficult to meet the application of high real-time requirements.On the basis of analysis of the principle of SIFT algorithm,in view of the large numbers of extracted features,high-dimensional of the feature vector,we refined the algorithm for parallel processing to take advantage of modern graphics hardware,and compared it with the CPU SIFT algorithms．Experiments demonstrate that the algorithm based on GPU SIFT significantly increases efficiency,and can reach the speed ratio of more than ten times averagely.

Key words: GPU,SIFT,CUDA,Feature matching

姜超,耿则勋,娄博,魏小峰,沈忱. 基于GPU的SIFT特征匹配算法并行处理研究[J]. 计算机科学, 2013, 40(12): 295-297. https://doi.org/

JIANG Chao,GENG Ze-xun,LOU Bo,WEI Xiao-feng and SHEN Chen. Parallel Processing Research on SIFT Feature Matching Algorithm Based on GPU[J]. Computer Science, 2013, 40(12): 295-297. https://doi.org/

参考文献

[1] 王瑞,梁华,蔡宣平.基于GPU的SIFT特征提取算法研究[J].现代电子技术,2010(15):41-43
[2] 盖素丽.基于GPU的数字图像并行处理研究[J].计算机技术理论,2009(2):38-40
[3] 张舒,楮艳利,赵开勇,等.GPU高性能运算之CUDA[M].北京:中国水利水电出版社,2009:1-10
[4] Lowe D G.Distinctive Image Features from Scale-Invariant Keypoints[J]．International Journal of Computer Vision,2004,60(2):91-110
[5] 洪宇,龚建华,胡社荣,等.无人机遥感影像获取及后续处理探讨[J].遥感技术与应用,2008,23(4):462-466
[6] 杨化超.宽基线立体影像点、线特征提取与匹配方法研究[D].徐州:中国矿业大学,2010:55-60
[7] Jan K.The structure of images [J].Biological Cybernetics,1984,50:363-370
[8] Lindeberg T．Scale-space theory:A basic tool for analysingstructures at different scales[J]．Journal of Applied Statistics,1994,21(2):224-270
[9] Sinha S N,Frahm J-M,Pollefeys M,et al．GPU-Based VideoFeature Tracking and Matching[R]．EDGE 2006,Workshop on Edge Computing Using New Commodity Architectures．Chapel Hill,2006:1-15
[10] Vedaldi A.sift++.http://vision.ucla.edu/~vedaldi/code/siftpp/siftpp.html
[11] Sudipta N S,Jan-Michael F,Marc P.Feature tracking and ma-tching in video using programmable graphics hardware[J]．Machine Vision & Applications,2011,2(1):207-217

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed