Local Affine Frames for Image Retrieval

A novel approach to content-based image retrieval is presented. The method supports recognition of objects under a very wide range of viewing and illumination conditions and is robust to occlusion and background clutter. Starting from robustly detected 'distinguished regions' of data dependent shape, local affine frames are established by affine-invariant constructions exploiting invariant properties of the second moment matrix and bi-tangent points. Direct comparison of photometrically normalised colour intensities in normalised frames facilitates robust, affine and illumination invariant, but still very selective matching. The potential of the proposed approach is experimentally verified on FOCUS -- a publicly available image database - using a standard set of query images. The results obtained are superior to the state of the art. The method operates successfully on images with complex background, where the sought object covers only a fraction (around 2%) of the database image. Examples of precise localisation of the query objects in an image are shown too.

[1]  Sang Uk Lee,et al.  Color image retrieval using hybrid graph representation , 1999, Image Vis. Comput..

[2]  Jiri Matas,et al.  Distinguished Regions for Wide-baseline Stereo , 2001 .

[3]  Marcel Worring,et al.  Content-Based Image Retrieval at the End of the Early Years , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  Jan Flusser,et al.  Convex Layers: A New Tool for Recognition of Projectively Deformed Point Sets , 1999, CAIP.

[5]  Michael J. Swain,et al.  Color indexing , 1991, International Journal of Computer Vision.

[6]  Fang Liu,et al.  Periodicity, Directionality, and Randomness: Wold Features for Image Modeling and Retrieval , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  L. Guibas,et al.  Finding color and shape patterns in images , 1999 .

[8]  Brian V. Funt,et al.  Color Angular Indexing , 1996, ECCV.

[9]  Alberto Del Bimbo,et al.  Effective image retrieval using deformable templates , 1996, Proceedings of 13th International Conference on Pattern Recognition.

[10]  Andrew Zisserman,et al.  Geometric invariance in computer vision , 1992 .

[11]  Luc Van Gool,et al.  Recognizing color patterns irrespective of viewpoint and illumination , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[12]  Brian V. Funt,et al.  Color Constant Color Indexing , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  Josef Kittler,et al.  Robust and Efficient Shape Indexing through Curvature Scale Space , 1996, BMVC.

[14]  Luc Van Gool,et al.  Content-Based Image Retrieval Based on Local Affinely Invariant Regions , 1999, VISUAL.

[15]  Rajiv Mehrotra,et al.  Shape-similarity-based retrieval in image database systems , 1992, Electronic Imaging.