A feature-based method for tire pattern similarity detection

Abstract

Innovation and inheritability are the key requirements of tire pattern design. This article call attention to the similarity detection of tire patterns. A similarity detection method based on tire pattern features is proposed. Through in-depth analysis of the pattern structure, the tire pattern features are separated into shape and texture features. Shape features describe the arrangement rules between tire grooves and pattern blocks. On the other hand, texture features define each pattern block’s surface characteristics, groove depth, groove width, and other characteristics. The tire pattern raw data is firstly analyzed through digital image processing technology. Then, shape and texture features were recognized and separated using edge extraction and co-occurrence matrix methods, respectively. The tire pattern photos are then realized, measured, and detected for similarity following the rules of similarity theory. Experiments were designed to fit the framework of the given detection method. The test results prove that the proposed method can detect tire pattern similarity quickly and conveniently.