Surface Image Feature Extraction of Laminated Flooring Based on Texture Analysis

Extracting surface image features of defective laminated flooring correctly and efficiently is the base of detecting its product quality which makes use of machine vision technology. Through the method of texture analysis, this article applies the Gray Level Co-occurrence Matrix (GLCM) to extract image features from a particular series of multiple defective images of laminated flooring. We acquire four principal eigenvalues that describe image texture, namely the energy, contrast, relevance, entropy. The experimental data show that the four eigenvalues can better reflect and express the main defects of laminated flooring.