Classification of Moroccan Decorative Patterns Based on Machine Learning Algorithms

Moroccan craftsmanship has been booming in recent years. The decorations mark refined traces of patterns and forms rich in colors. Despite recent studies, the machine still finds problems recognizing patterns and shapes. In this paper, we propose a new method of classifying decor patterns based on the analysis of the texture and color of the images. Texture information describes the cross section of the pattern while color describes the distribution of colors in the image. In the majority of cases, the Moroccan decoration is composed of five basic elements: ceramic, plaster, fabric, glass, and wood. In a first step, we based ourselves on the techniques of extraction of the characteristics of texture and color. The techniques we used for the description of the images are the Gray-Level Co-occurrence Matrix (GLCM) and Local Binary Patterns (LBP) combined with the Color Histogram (CH). In a second step, we proposed a machine learning classification model based on K-Nearest Neighbors (K-NN) and Support Vector Machine (SVM). The results obtained for the classification of set images exceed an accuracy of 86%.