This paper studies the automatic test system of bus dashboard EOL (end of line) based on machine vision. Based on machine vision theory, Identification and detection algorithm of panel signal indicator elements and tachometer pointer readings was studied combining single-frame still images and real-time processing of color video image, the automatic parallel detection of multiple dashboard was realized by distributed network architecture. This paper first describes the function requirements, the overall composition and working principle of automatic test system. Then, it proposes an automatic identification and detection algorithm of dashboard symbol sheets and pointer position. Finally, it shows the designing of automatic test software with a self-learning and auto-detection function, and describes the working process of the software. The tests prove that the system is capable of realizing fast and accurate auto-test of bus dashboard functions based on the non-contact of machine vision, which improves the overall efficiency of the bus dashboard line.
Automatic test system of the back-end card for the JUNO experiment
