Research on Machine Vision Detection Experiment Platform for Industry Products

摘要

Key technologies of machine vision detection system for the industrial part are investigated and an intact set of high-speed detection system is developed. On the basis of introducing the entire design of machine vision platform, analyzing the software and hardware structure and fundamental principles of control units, this paper proposes that variance function shall be adopted for coarse tuning and the search algorithm of Roberts gradient sum function shall be adopted for fine tuning. The experimental results show that it only takes 3.246 second for this system to collect images, which meets the requirement of real-time online detection. This system is feasible for the detection experiments of industrial products.