System and method for automatic treadwear classification

Abstract

A system and method are provided for automatically alerting drivers to potential tread ware problems to enable them to avoid the danger hazards associated with worn treads. A tread-evaluation station is placed at a location where images of tires may be captured. Images of tires are recorded when a vehicle is at or near the tread-evaluation station. An automated analysis is performed on the images. Based on the automated analysis, tires depicted in the captured images are classified into categories of wear. The automated analysis may include a detection component trained to detect tires in images, and a classification model trained to assign a classification to the wear status of the tires identified by the detection component. The tread-evaluation station may further be trained to predict when tires that do not currently need replacing will need replacing.