Motion detection algorithms using template model

Abstract

The Horridge Template model is an empirical motion detection model inspired by insect vision. This model has been successfully implemented in several micro-sensor VLSI chips using grayscale pixels. The template model is based on movement of detected edges rather than the whole object, which consequently facilitates simple tracking techniques. Simple tracking algorithms developed by Nguyen have been successful in tracking coherent movement of objects in a simple environment. An extension of the template model using color templates developed by Chin has also been successful in tracking objects moving in close proximity to each other. This paper introduces a low-cost insect vision prototype based on the use of a color CMOS camera. We implement a further extension to the above algorithms with error checking. Several error-checking schemes are used during template formation and manipulation to reduce noise and randomness. This enables the detection of moving objects in noisy environments, which may be applied to many real life situations.