Wear Debris, Classification, Supervised Machine Learning, SVM, Linear Regression,

Features, PCA, HOG, LBP

The wear debris of various engineering equipment (such as combustion engines, gearboxes, etc.) consists of particles of metal which can be obtained from lubricants used in such machine parts. The analysis the wear particles is very important for early detection and prevention of failures in engineering equipment. The analysis is often done through classification of individual wear particles obtained by analytical ferrography. In this paper, we present a study of feature extraction methods for a classification of the wear particles based on visual similarity (using supervised machine learning). The main contribution of the paper is the comparison of nine selected feature types in the context of three state-of-the-art learning models. Another contribution is the large public database of binary images of particles which can be used for further experiments.