A Biologically Motivated Corner Detection Method Based on the Oriented Receptive Fields of Simple Cortical Cells

Abstract

Corner detection is an important early vision problem. In early vision processing of the mammalian visual system, the oriented tuned, contrast-driven cells in the visual cortex have the problems of positional and orientational uncertainty. That means the oriented receptive fields (RFs) of simple cells can not detect the orientational information of the line ends and corners efficiently. In this paper, we use the uncertainty property of the RFs to detect corners in gray-level images, where Gabor functions are used to model the RFs of simple cells. The proposed method is proved to be efficient for patches of texture, uneven surfaces and geometric discontinuities. Experimental results with some synthetic and natural images show that this new method has good performances and is robust to noise.