Vision-Based Inspection of Tyre Tread Depth

Abstract

In this paper, an approach for visual, non-contact automatic inspection of tyre tread depth based on existing image processing techniques is presented. Histograms of oriented gradient are used for feature extraction from images. In order to analyse which set of features gives the best classification results, a linear support-vector machine classifier was trained and tested using different numbers of pixels and numbers of cells per block. The obtained processing and experimental results are presented in this paper.