Aiming at the problems of low efficiency and easy introduction of errors in visual readings when manually verification pressure meter,an automatic verification system of pressure meter based on machine vision was designed,and the control software and recognition method were developed.The control software was designed for automation processes,introducing multi-threading,containerization,and monitoring event technology.It controlled hardware such as PLC,robotic arm,industrial camera,and standard pressure generator in the system to achieve automation of the calibration process,and supported simultaneous verification of multiple pressure meter.Considering the low configuration of industrial computers,a lightweight recognition method was designed based on the deep learning model Paddle and the OpenCV library.This method not only takes up less memory and has fast operation speed,but also has richer recognition information and higher recognition accuracy.The experimental results show that the system can effectively improve verification,reduce human error,and has application and promotion value.