Expert computer vision based crab recognition system

Two species of crabs are Chionoecetes bairdi and C. opilio and their hybrids live in the Bering Sea. The two species differ in generic, morphological and morphometric characteristics. Inter breeding of C. bairdi and C. opilio results in a hybrid form with intermediate morphological and morphometric characteristics. These species were determined by analyzing the empirical covariance matrices associated with the two species and the hybrid and by taking into account the statistical reliability of the estimated principal components. A modified eigen image classifier is implemented based on the assumption that the data is drawn from multivariate Gaussian distributions with different means and covariance matrices. The clustering technique is introduced to minimize the misclassification rate of the eigen image classifier