Research on Machine Vision Technology of High Speed Robot Sorting System Based on Deep Learning

With the improvement of industrial automation level and the progress of science and technology, the number of robots is increasing, the application scenarios are becoming more and more complex, and the requirements for automation, intelligence, precision, stability and flexibility of robots are also increasing. Machine vision refers to the use of machines instead of human eyes for measurement and judgment. The ultimate goal of machine vision is to enable machines to observe and understand the input image data accurately like human eyes, and finally make decisions to achieve the purpose of adapting to the environment autonomously. In the traditional industrial production line, the task of sorting workpieces is carried out manually, which is not only inefficient but also costly. It is the trend of industrial automation to apply machine vision technology to sorting tasks of industrial robots. According to the actual needs of China's production industry, based on advanced technologies such as deep learning algorithm and machine vision, this paper constructs a high-speed robot sorting system for product production to improve the overall operation effect of the robot sorting system.