Fast Inspection of Food Packing Seals Using Machine Vision

This paper presents a machine vision approach to inspecting the integrity of food packing seals. A kind of new method for inspecting food packing seals was used, the method meets currently the industry's requirements for high-speed and accuracy quality inspection, with low associated production costs. Through analyzing examples of “eligible” and “faulty” food packing seals images, these seals uniformity have been examined and associated with numerical quality measures. For distinguishing between “eligible” and “faulty” seals conditions, template matching is used. Getting R_{(i, j)} correlation parameter between template and image inspected, then it was taken as standard distinguishing “eligible” and “faulty” seals. For realizing fast inspection, a set of inspecting and grading system based on machine vision was developed. The system is shown to have a high success rate provided that illumination conditions remain constant.