Abstract

The compound eyes of insect are highly optimized for the task of visual acquisition of relevant information from the environment, such as moving object detection. These have attracted the attention of many researchers in recent years. The advantages of compound eyes are illustrated in this paper in relation to other approaches in moving object detection. The paper describes the imaging mechanism for the compound eyes, and we also describe the process of multi-resolution imaging. An image filter with variable resolution is discussed and we propose a new method for moving object detection using variable resolution double-level contours. Firstly, there is the contour detection which occurs in the lowest resolution images. When this is complete, texture gradient computation is performed which carries out contour segmentation in the high resolution image acquired by the single eye. Finally, a Geodesic Active Contour model for texture boundary detection is applied from which the moving target is recognized. The results in this paper show that this method has good potential as a solution for moving object detection problems.

Key words: Moving object detection, bionic compound eyes, variable resolution, double-level contours.