Inspect species recognition featuring projection-based shape descriptor

Abstract

This paper proposes an automated system for recognizing insect species based on their shape. In this report, a Radon Transform shape descriptor is proposed to extract a projection according to a single angle. This projection is chosen in way that contains the necessary information to recognize an object. This descriptor (called RƟ-signature) provides global information of a binary shape regardless its form. The signature keeps fundamental geometrical transformation like scale, translation and rotation. Implemented insect recognition sample tested in used software environment and the test results were analyzed. The accuracy and the efficiency of the proposed algorithm in the presence of a variety of transformations are evaluated within a shape recognition process. This project will add an assistant feature of Neural Network Classifier tool, and this feature will effectively enhance the efficiency of retrieval in largest dataset.