Local Features and Kernels for Classication of Texture and Object Categories: A Comprehensive Study
Jianguo Zhang, Svetlana Lazebnik, Cordelia Schmid · 2006
Recently, methods based on local image features have shown promise for texture and object recognition tasks. This paper presents a large-scale evaluation of an approach that represents images as distributions (signatures or histograms) of features extracted from a sparse set of keypoint locations and learns a Support Vector Machine classier with kernels based on two effective measures for comparing distributions, the Earth Mover’s Distance and the 2 distance. We r st evaluate the performance of our approach with different keypoint detectors and descriptors, as well as different kernels and classier s. We then conduct a comparative evaluation with several state-of-the-art recognition methods on 4 texture and 5 object databases. On most of these databases, our implementation exceeds the best reported results and achieves comparable performance on the rest. Finally, we investigate the inuence of background correlations on recognition performance.