Saliency map model with adaptive masking based on independent component analysis

Abstract We propose a new saliency map model with an adaptive masking method. Independent component analysis (ICA) is used to find a filter that can generate a saliency point from a color natural scene. Also, we consider the roles of cells in retina such as edge detection and color opponency as a preprocessor of an ICA filter that models the brain function to reduce the redundancy. In order to adopt a suitable masking in a salient point, a noise-tolerant generalized symmetry transformation is incorporated into the morphology operation. Computer experimental results show that the proposed model successfully generates the sequence of salient points.