An automated visual inspection system for the glass industry

Today, computer vision inspection systems are widely used for quality control in order to reduce the costs and to improve the product quality. The glass industry is constantly trying to improve quality by substituting the human control with automated inspection systems but several problems must be solved. In this paper the problem of detecting and measuring the defects of satin glasses is investigated and a real-time system is proposed that is able to analyze the glass surface under inspection, to assess its quality and to characterize its defects. A prototype has been carefully designed and optimized for validating the proposed approach and to reproduce the real issues of quality control. The prototype is composed by several CMOS cameras, a controllable conveyor bend, and an image processing system. Currently the prototype, which is cheap and reliable, is under further development in cooperation with a specialized electronic industry.