Identification of the American burying beetle, Nicrophorus americanus, through image processing

Abstract: The American burying beetle (ABB) (Nicrophorus americanus) is a critically endangered species that has declined to 10% of its original range. Current ABB trapping methods are labor intensive and provide limited temporal data. The objective of this study is to develop an autonomous monitoring system for existing ABB traps that will detect insects as they enter a trap and classify them as ABB or other species. Various morphological features, including size, shape and color components, were extracted from captured images, and experiments using several classifiers were performed. Template matching recognition protocols on 11 species of insects from a museum collection resulted in correct ABB recognition of 80% (n=10) whereas overall classification accuracy for all 11 species evaluated was 75% (S.D. 13%). Further research will be conducted on improving classification results of ABB with color image processing.