OpenCV-based SF6 density controller dial pointer angle recognition
Zhangsi Yu, Ning Zhang, Hou‐Cheng Yang · Journal of Physics Conference Series · 2023
Abstract To solve the problems of the low efficiency and poor accuracy of the manual detection in the factory inspection of the SF6 density controller, this paper proposes an SF6 density controller dial pointer angle identification method. First, the image of the dial pointer at a specific position is acquired according to the working characteristics of the SF6 density controller. The center and radius of the contour circle inside the scale line in the dial are extracted using the Hoff gradient method and the image is coordinate transformed using this parameter. Then the coordinate-transformed image is pre-processed. And finally, the threshold of the image after pre-processing is traversed and the length is used to characterize the pointer angle through the correspondence between the scale line and the angle. The angle of the pointer is characterized by the correspondence between the scale and the angle using the length. This paper uses Python and OpenCV vision library to complete the algorithm and through experiments, the algorithm can achieve high precision pointer angle recognition.