Anatomy and Physiology of an Artificial Vision Matrix

Abstract

Abstract. We present a detailed account of the processing that occurs within a biologically-inspired model for visual homing. The Corner Gradient Snapshot Model (CGSM) initially presented in [1] was inspired by the snapshot model [2] which provided an algorithmic explanation for the ability of honeybees to return to a place of interest after being displaced. The concept of cellular vision is introduced as a constraint on processing. A cellular vision matrix processes visual information using retinotopically arranged layers of low-level processing elements interacting locally. This style of processing reflects general principles known of visual processing throughout the animal kingdom. From a technical standpoint, this style of processing is inherently parallel. Here we describe a cellular vision matrix which implements CGSM and illustrate how this matrix obeys cellular vision. Some new comparative results are presented and it is found that CGSM's performance degrades gracefully with environmental modification and occlusion.

A Scale Invariant Local Image Descriptor for Visual Homing

“…Both camps have proposed methods which find correspondences between image features and use these to compute a home vector. These featurebased methods rely on visual features such as regions in 1-D (one-dimensional) images [4,5], edges in 1-D images [3], image windows around distinctive points in 1-D images [2], coloured regions in 2-D images [6], and Harris corners in 2-D images [7,8]. Any visual feature is subject to distortions in scale, illumination, and perspective, as well as distortions from occlusion.…”