Machine vision is a rapidly developing branch of artificial intelligence, which uses machines to simulate human visual functions. Its working principle is that the captured objects are converted into image signals through machine vision products (including CMOS and CCD), and then transmitted to a special image processing system for various operations to extract information and understand it, and finally used for actual identification, detection, measurement and control.