Improved Defect Detection Using Novel Wavelet Feature Extraction Involving Principal Component Analysis and Neural Network Techniques.

Abstract

This paper aims at investigating a novel solution to the problem of defect detection from images, that can find applications in the design of robust quality control systems for the production of furniture, textile, integrated circuits, etc. The suggested solution focuses on detecting defects from their wavelet transformation and vector quantization related properties of the associated wavelet coefficients. More specifically, a novel methodology is investigated for discriminating defects by applying a supervised neural classification technique, employing a Multilayer Perceptron (MLP) trained with the conjugate gradients algorithm, to innovative multidimensional wavelet based feature vectors. These vectors are extracted from the K-Level 2-D DWT (Discrete Wavelet Transform) transformed original image using Vector Quantization techniques and a Principal Component Analysis (PCA) applied to these wavelet domain quantization vectors. The results of the proposed methodology are illustrated in defective textile images where the defective areas are recognized with higher accuracy than the one obtained by applying two rival feature extraction methodologies. The first one of them uses all the wavelet coefficients derived from the k-Level 2-D DWT, while the second one uses only image intensities characteristics. Both rival methods involve the same classification stage as the proposed feature extraction approach. The promising results herein obtained outline the importance of judicious selection and processing of 2-D DWT wavelet coefficients for industrial pattern recognition applications.