Fast codebook generation by sequential data analysis for object classification

Abstract

In this work, we present a novel, fast clustering scheme for codebook generation from local features for object class recognition. It relies on a sequential data analysis and creates compact clusters with low variance. We compare our algorithm to other commonly used algorithms with respect to cluster statistics and classification performance. It turns out that our algorithm is the fastest for codebook generation, without loss in classification performance, when using the right matching scheme. In this context, we propose a well suited matching scheme for assigning data entries to cluster centers based on the sigmoid function. © Springer-Verlag Berlin Heidelberg 2007.