Aiming at the problem that when the optical axis of the acquisition device deviates from the center of the meter

the projected position of the pointer on the pointer-type meter in the image will deviate from the actual position

making it difficult to accurately recognize the reading

an identification and interpretation algorithm for pointer-type meters based on binocular stereoscopic vision is proposed. Firstly

the image of pointer-type meter is captured by using calibrated binocular camera

and the left and right images are aligned in the horizontal direction by stereoscopic correction. Then

the pixel coordinates of the pointer tip and the target points on the meter panel plane in the left image can be recognized based on threshold segmentation and Hough transform

which determines the initial position of the pointer tip. And similarly

the gray matching algorithm is used to extract the corresponding points in the right image. Finally

the three-dimensional coordinates of the target points are calculated by the distance measurement based on binocular stereoscopic vision and the meter plane can be constructed with the three-dimensional information. The pointer tip is projected vertically on the plane to reposition the pointer tip. The recognized reading can be output. The experimental results show that the algorithm has strong stability and applicability. Compared to the results of pointer reading recognition algorithm based on the Hough transform

this algorithm in this paper reduces the error by approximately 64 %.