Configurable image-processing system for education in automotive applications

This paper describes the overall image-processing system, which has been developed for the use in education and at various industrial projects. The added value for the students is the tight link between computer-based education in image processing and the real-world relevance. Complex and changing quality control demands of safety critical components for automotive applications require an innovative graphical systems design, which can be effortlessly reconfigured by the user. The quality control challenge claimed us to detect imperfections on the backside of brake shoes. These imperfections can cause safety related problems and therefore, the affected brake shoes need to be sorted out. From an engineering point of view, this was a technically challenging image-processing application. The surface of the brake shoe is highly reflective; the contrast between the imperfections and remaining areas is very low, the selection criteria are complex (depending on the size, number, and location of the imperfection). Due the fact that every car manufacturer and almost every car model needs different brake shoes, the most demanding requirement was, that the shape of the brake shoe needs to be freely configurable by the user.