The Community for Technology Leaders
Green Image
Issue No. 09 - Sept. (2012 vol. 34)
ISSN: 0162-8828
pp: 1758-1772
A. Y. Chia , Institutefor Infocomm Res., Singapore, Singapore
D. Rajan , Sch. of Comput. Eng., Nanyang Technol. Univ., Singapore, Singapore
M. K. Leung , Dept. of Comput. Sci., Univ. Tunku Abdul Rahman, Kampar, Malaysia
S. Rahardja , Institutefor Infocomm Res., Singapore, Singapore
We present a novel contour-based approach that recognizes object classes in real-world scenes using simple and generic shape primitives of line segments and ellipses. Compared to commonly used contour fragment features, these primitives support more efficient representation since their storage requirements are independent of object size. Additionally, these primitives are readily described by their geometrical properties and hence afford very efficient feature comparison. We pair these primitives as shape-tokens and learn discriminative combinations of shape-tokens. Here, we allow each combination to have a variable number of shape-tokens. This, coupled with the generic nature of primitives, enables a variety of class-specific shape structures to be learned. Building on the contour-based method, we propose a new hybrid recognition method that combines shape and appearance features. Each discriminative combination can vary in the number and the types of features, where these two degrees of variability empower the hybrid method with even more flexibility and discriminative potential. We evaluate our methods across a large number of challenging classes, and obtain very competitive results against other methods. These results show the proposed shape primitives are indeed sufficiently powerful to recognize object classes in complex real-world scenes.
shapes (structures), image classification, image representation, object detection, object recognition, object detection, line segment, ellipse, appearance feature, contour-based approach, object classes recognition, generic shape primitive, representation efficiency, storage requirement, geometrical property, shape-token, shape structure, hybrid recognition method, image classification, Shape, Feature extraction, Training, Image edge detection, Vectors, Robustness, Image segmentation, category-level object detection., Shape primitives, appearance features, image classification
A. Y. Chia, D. Rajan, M. K. Leung, S. Rahardja, "Object Recognition by Discriminative Combinations of Line Segments, Ellipses, and Appearance Features", IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 34, no. , pp. 1758-1772, Sept. 2012, doi:10.1109/TPAMI.2011.220
159 ms
(Ver 3.3 (11022016))