Defect detection in textiles using special techniques fortexture analysis

Abstract

This paper presents the use of spatial analysis techniques texture images for defects identification in textiles, where the application of pre-processing techniques is presented, and the analysis of texture local binary patterns and co-occurrence matrix to extract features to highlight differences and identify defects in textiles using classifier based on the Mahalanobis distance.

Keywords

• análisis de textura, distancia de mahalanobis, defectos en textiles, matriz de co-ocurrencia, patrones locales binarios, filtrado de imágenes.