Neural Network Based Classification Using Blur Degradation and Affine Deformation Invariant Features

Identification of affine deformed and simultaneously blur degraded images is an important task in pattern analysis. Use of global moment features has been one of the most popular techniques for pattern recognition and classification. In this paper, we introduce an approach to derive blur and affine combined moment invariants(BACIs). A neural network(NN) model is then employed to classify objects using these BACIs. Introduction The objective of a typical computer vision system is to analyze images of a given scene and recognize the content of the scene. Most of these systems share a general structure which is composed of four building blocks: image acquisition, preprocessing, feature extraction, and classification. The main focus of this paper is on the feature extraction and classification problems. hnages to be processed are usually unsatisfactory with geometric distortion and/or blur degradation. About geometric deformations, we mainly discuss the 2-D general affine transformation, which transforms the original image f(x, y) to a new image f’(x ~, y~) and ha:~ the following form: