Fourier Transform and Image Processing in Automated Fabric Defect Inspection System

Automated fabric inspection system is important to prevent delivering of inferior quality fabric and designed to increase the accuracy, consistency and speed of defect detection in fabric manufacturing process to reduce labor costs, improve product quality and increase manufacturing efficiency. Fabric inspection is still carried out offline and manually by humans with many drawbacks such as tiredness, boredom and inattentiveness. The continuous development in computer technology introduces the online fabric inspection system based on image processing as an effective alternative of offline inspection system. This paper proposes an effective and accurate approach to automatic fabric inspection system. The defect free fabric has a periodic regular structure, the occurrence of a defect in the fabric breaks the regular structure. Therefore the fabric defects can be detected by monitoring fabric structure. The Fabric inspection system first acquires high resolution, high contrast and minimum noise of image with suitable format. In this paper Fast Fourier transform technique and cross correlation techniques are implemented on plain fabric to examine the structure regularity features of the image. Keywords-Cross Correlation technique, Fabric Inspection, Fourier Transform, Image Processing.