计算机科学 ›› 2016, Vol. 43 ›› Issue (Z6): 165-167.doi: 10.11896/j.issn.1002-137X.2016.6A.038

• 模式识别与图像处理 • 上一篇    下一篇

移动智能终端的SIFT特征检测并行算法

甘威,张素文,雷震,李怡凡   

  1. 武汉理工大学自动化学院 武汉430070,武汉理工大学自动化学院 武汉430070,武汉理工大学自动化学院 武汉430070,武汉理工大学自动化学院 武汉430070
  • 出版日期:2018-12-01 发布日期:2018-12-01

SIFT Feature Extraction Parallel Algorithm on Mobile Device

GAN Wei, ZHANG Su-wen, LEI Zhen and LI Yi-fan   

  • Online:2018-12-01 Published:2018-12-01

摘要: 特征的检测和匹配在计算机视觉应用中是一个重要的组成部分,如图像匹配、物体识别和视频跟踪等。SIFT算法以其尺度不变性和旋转不变性在图像配准领域得到了广泛应用。传统的SIFT算法效率低,因此提出一种在移动智能终端上实现的高效方法。在Android平台利用OpenCL框架实现了移动智能终端的SIFT算法,通过计算任务的重新分配,优化SIFT算法在移动GPU上的并行实现。实验结果表明,移动平台的SIFT算法充分利用了GPU并行计算能力,大大提高了SIFT算法的执行效率,实现了高效的特征检测。

关键词: SIFT,OpenCL,GPU,特征检测

Abstract: Feature extraction and matching is an important part in computervision applications,such as image matching,object recognition and video tracking.SIFT algorithm is widely used in the field of image registration because of its scale invariance and rotation invariance.To deal with the low efficiency of the traditional SIFT algorithm,we proposed an efficient method which is implemented on mobile platform.In this paper,we used the OpenCL to achieve the SIFT algorithm on mobile device,through redistributing the calculation of tasks and optimizing the SIFT algorithm in the mobile OpenCL parallel implementation.Experimental results show that our SIFT algorithm takes full advantage of the GPU parallel computing power,greatly improving the efficiency of the SIFT algorithm,achieving the efficient feature extraction.

Key words: SIFT,OpenCL,GPU,Feature extraction

[1] Lowe D G.Distinctive Image Features from Scale-Invariant Keypoints[J].International Journal of Computer Vision,2004,60(2):91-110
[2] Sinha S N,Frahm J M,Pollefeys M,et al.Feature tracking and matching in video using programmable graphics hardware[J].Machine Vision & Applications,2011,22(1):207-217
[3] Kayombya G R.SIFT feature extraction on a Smartphone GPU using OpenGL ES2.0[D].Massachusetts:Massachusetts Institute of Technology,2011
[4] Rister B,Wang G,Wu M,et al.A fast and efficient sift detector using the mobile GPU[C]∥2013 IEEE International Confe-rence on Acoustics,Speech and Signal Processing (ICASSP).IEEE,2013:2674-2678
[5] Wang Wei-yan,Zhang Yun-quan,Long Guo-ping,et al.CLSIFT:An Optimization Study of the Scale Invariant Feature Transform on GPUs[C]∥IEEE 15th International Conference on High Performance Computing and Communication.2013
[6] Wang G,Rister B,Cavallaro J R.Workload Analysis and Efficient OpenCL-based Implementation of SIFT Algorithm on a Smartphone[C]∥ Global Conference on Signal and Information Processing(GlobalSIP).IEEE,2013:759-762
[7] Wang G.An easy-to use standalone SIFT library easy-to-usestandalone SIFT library written in C/C++[CP/OL].http://sourceforge.net/projects/ezsift
[8] 张樱,张云泉,龙国平.基于OpenCL的图像模糊化算法优化研究[J].计算机科学,2012,39(3):260-264
[9] 贾海鹏,张云泉,徐建良.基于OpenCL的图像积分图算法优化研究[J].计算机科学,2013,0(2):1-7
[10] 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,2010,33(15):41-43

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!