Color-based scale-invariant feature detection applied in robot vision

The scale-invariant feature detecting methods always require a lot of computation yet sometimes still fail to meet the real-time demands in robot vision fields. To solve the problem, a quick method for detecting interest points is presented. To decrease the computation time, the detector selects as interest points those whose scale normalized Laplacian values are the local extrema in the nonholonomic pyramid scale space. The descriptor is built with several subregions, whose width is proportional to the scale factor, and the coordinates of the descriptor are rotated in relation to the interest point orientation just like the SIFT descriptor. The eigenvector is computed in the original color image and the mean values of the normalized color g and b in each subregion are chosen to be the factors of the eigenvector. Compared with the SIFT descriptor, this descriptor's dimension has been reduced evidently, which can simplify the point matching process. The performance of the method is analyzed in theory in this paper and the experimental results have certified its validity too.

[1]  Cordelia Schmid,et al.  Comparing and evaluating interest points , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[2]  Diego Viejo,et al.  Active stereo based compact mapping , 2005, 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[3]  J. Crowley.  A representation for visual information , 1981 .

[4]  Glenn Healey,et al.  The Illumination-Invariant Recognition of 3D Objects Using Local Color Invariants , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Gustavo Carneiro,et al.  Multi-scale phase-based local features , 2003, 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings..

[6]  Cordelia Schmid,et al.  A Performance Evaluation of Local Descriptors , 2005, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  James J. Little,et al.  Vision-based global localization and mapping for mobile robots , 2005, IEEE Transactions on Robotics.

[8]  Xinde Li,et al.  A Quick Feature Detecting Method Applied in Robot Vision , 2007, 2007 International Conference on Mechatronics and Automation.

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

[10]  C. Schmid,et al.  Indexing based on scale invariant interest points , 2001, Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001.

[11]  Cordelia Schmid,et al.  Local Grayvalue Invariants for Image Retrieval , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[12]  Tony Lindeberg,et al.  Feature Detection with Automatic Scale Selection , 1998, International Journal of Computer Vision.

[13]  Illah R. Nourbakhsh,et al.  Appearance-based place recognition for topological localization , 2000, Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No.00CH37065).

[14]  Hongbin Zha,et al.  Coarse-to-fine vision-based localization by indexing scale-Invariant features , 2006, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).