Footprint Identification of Weta and Other Insects

Abstract

Using tracking tunnels or tracking cards to obtain the footprints of active animals living in a specific area is useful for biosecurity research as well as environment monitoring. In the absence of human experts, a feasible computer classifier could be constructed to identify species of insects from their footprints. The essential step to develop this classifier is to obtain representative attributes on which we can perform analysis using image processing technology, geometry, etc. This paper presents procedures for an approach to build up such a classifier by discussing three different attribute heuristics.