The use of color in wear debris analysis

Abstract

A method and results of classification of metallic wear debris using their color features are presented. The color images of wear debris were used as initial data. The color features of the debris are specified by an HSI color model, Each particle is characterized by a set of statistical features derived from a distribution of HSI color model components. The initial feature set was optimized by principal component analysis. A multidimensional scaling procedure was used for definition of the classification plane. The selected features allow us to distinguish copper-based alloys, red and dark iron oxides and steel particles. A method was proposed of probabilistic decision of class assignment based on analysis of debris coordinates distribution in the classification plane. The obtained results have demonstrated the feasibility of using color features for automated wear particle analysis. (C) 2001 Elsevier Science B.V. All rights reserved.