GPU-ASIFT: A fast fully affine-invariant feature extraction algorithm

This paper presents a method that takes advantage of powerful graphics hardware to obtain fully affine-invariant image feature detection and matching. The chosen approach is the accurate, but also very computationally expensive, ASIFT algorithm. We have created a CUDA version of this algorithm that is up to 70 times faster than the original implementation, while keeping the algorithm's accuracy close to that of ASIFT. It's matching performance is therefore much better than that of other non-fully affine-invariant algorithms. Also, this approach was adapted to fit the multi-GPU paradigm in order to assess the acceleration potential from modern GPU clusters.

[1]  Luo Juan,et al.  A comparison of SIFT, PCA-SIFT and SURF , 2009 .

[2]  Jean-Michel Morel,et al.  ASIFT: A New Framework for Fully Affine Invariant Image Comparison , 2009, SIAM J. Imaging Sci..

[3]  Jan-Michael Frahm,et al.  Feature tracking and matching in video using programmable graphics hardware , 2007, Machine Vision and Applications.

[4]  Derek Hoiem,et al.  Pascal VOC 2008 Challenge , 2008 .

[5]  Thomas Wiegand,et al.  SIFT Implementation and Optimization for General-Purpose GPU , 2007 .

[6]  J. Morel,et al.  Is SIFT scale invariant , 2011 .

[7]  G. Griffin,et al.  Caltech-256 Object Category Dataset , 2007 .

[8]  Andreas Zell,et al.  Localization of mobile robots with omnidirectional vision using Particle Filter and iterative SIFT , 2006, Robotics Auton. Syst..

[9]  Victor Podlozhnyuk,et al.  Image Convolution with CUDA , 2007 .

[10]  Christopher Hunt,et al.  Notes on the OpenSURF Library , 2009 .

[11]  Yakup Genc,et al.  GPU-based Video Feature Tracking And Matching , 2006 .

[12]  James J. Little,et al.  Mobile Robot Localization and Mapping with Uncertainty using Scale-Invariant Visual Landmarks , 2002, Int. J. Robotics Res..

[13]  James J. Little,et al.  Vision-based mobile robot localization and mapping using scale-invariant features , 2001, Proceedings 2001 ICRA. IEEE International Conference on Robotics and Automation (Cat. No.01CH37164).

[14]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[15]  Amy W. Apon,et al.  Accelerating SIFT on parallel architectures , 2009, 2009 IEEE International Conference on Cluster Computing and Workshops.

[16]  Renaud Keriven,et al.  GPU-boosted online image matching , 2008, ICPR.

[17]  Matthew A. Brown,et al.  Recognising panoramas , 2003, Proceedings Ninth IEEE International Conference on Computer Vision.

[18]  G. Amdhal,et al.  Validity of the single processor approach to achieving large scale computing capabilities , 1967, AFIPS '67 (Spring).

[19]  M. Gokmen,et al.  Traffic sign recognition using Scale Invariant Feature Transform and color classification , 2008, 2008 23rd International Symposium on Computer and Information Sciences.

[20]  Changchang Wu,et al.  SiftGPU : A GPU Implementation of Scale Invariant Feature Transform (SIFT) , 2007 .

[21]  David G. Lowe,et al.  Shape indexing using approximate nearest-neighbour search in high-dimensional spaces , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[22]  S. Govindarajulu,et al.  A Comparison of SIFT, PCA-SIFT and SURF , 2012 .

[23]  Lionel Moisan,et al.  A Probabilistic Criterion to Detect Rigid Point Matches Between Two Images and Estimate the Fundamental Matrix , 2004, International Journal of Computer Vision.