Science intensive technology for ceramic tile quality control using computer vision

Abstract

This paper is a result generalization of theoretical investigations and a practical realization of the system of computer vision to control ceramic tile technological parameters. A basic problem is a choice of hardware and software for the definition and analysis of moving objects according to a sequence of images obtained during small time intervals. To identify similar objects against a background complex enough, but motionless it is necessary to define areas in which a motion is supposed. As a result of investigations of the areas found they may be changes up to dimensions of objects (that is, find objects themselves) and define parameters of their motion. At the same time, the number and dimensions of objects in the image can change within wide limits. The analysis carried out has shown that the application of OpenCV library allows simplifying the realization of such a system, and a developed programming module may be used in other fields of automated production.