A learning-based spatial processing method for the detection of point targets

Abstract

In this paper, we present an efficient learning-based method for the detection of point targets in images. In the scheme, the probabilistic visual learning (PVL) technique is used for modeling the appearance of point targets and constructing a saliency measure function. Based on this function and the feature vector extracted at each pixel position and a target saliency map is formed by lexicographically scanning the input image. We treat such saliency map as a spatially filtered result of input image. Experimental results show that the proposed algorithm outperforms other filter-based methods. © Springer-Verlag Berlin Heidelberg 2006.