A Novel Shape Indexing Method for Automatic Classification of Lepidoptera

This paper presents a novel approach for images retrieval form digital collection based on analysis of their low level characteristics such as color and shape that has been applied for classification of the rare specimens of the Lepidoptera (butterflies) in Mexico. The similarity with user's queries which are the images, manual sketches or textual descriptions is computed by comparison of the feature vectors representing a shape by proposed two segments turning functions that is invariant to translation, rotation, and scale. In order to speed up preprocessing of shapes the proposed convex regions algorithm and discrete curve evolution approach are applied. Another goal of the proposed method is integration of two signatures of shape used in hashing tables for data base of images such as compactness and elongatedness which provide fast and satisfactory image retrieval by accelerating the convergence to expected result. The proposed method has been tested and evaluated using designed search engine for classification of the butterfly's families