Species–identification of wasps using principal component associative memories

摘要

This paper presents a novel approach to image-based insect specimen identification, exploiting the ability of principal component auto associative memories to form trainable classifiers, which may be used to identify unknown images. The system utilises the differences between a pair of reconstructed images produced when the unknown image is included in, and then excluded from the training set encoded by the auto associative memory. A non-parametric statistical correlation metric, Kendall's, was used to correlate the reconstructed images. The approach has been applied to the species-identification of closely related parasitic wasps based upon their wing venation and pigmentation patterns.