Comparative study on classification of machined surfaces using ML techniques applied to GLCM based image features

The identification of the machined surface texture is very crucial in modern manufacturing industries. The surface texture analysis using machine vision, image processing, and classification using ML is a well-known domain of research in the last many years. This manuscript addresses the classification of machined surfaces (turned, ground, and shaped) using image processing and ML techniques. The machined surface images are captured using a DSLR camera, pre-processed, and partitioned into sixteen equal, nonoverlapping regions. The partitioned images are processed to extract the GLCM based features. The extracted features are fed to the ML classifiers such as decision tree, K-nearest neighbour, logistic regression, Naïve Bayes classifier, random forest, and support vector machine. All the ML techniques can be used for the classification of machined surface images. For this work, the random forest technique was found to provide the best performance in image classification.