An automatic reading method for a multi-dial pointer instrument

Abstract

The invention discloses an automatic reading method for a multi-dial pointer meter, which includes establishing reference images F of N dials included in the pointer meter bi , establish the images to be tested F of the N dials included in the pointer instrument ti , the image to be tested F of the dial i ti with reference image F bi Perform position comparison and calculate the reading of dial i p These three steps; the present invention sorts and numbers the images of a plurality of dials, and then extracts ROI, opening operation, pointer refinement, Hough transform line detection and meter scale calculation for each dial respectively, and automatically obtains the information on the pointer meter. The indication value of each dial breaks the limitation that an image can only identify a single instrument, saves image acquisition time and storage space, thereby improving image recognition efficiency. At the same time, the present invention uses morphological operations to reduce noise on pointer area images And through the fast image thinning algorithm, the pointer area image is thinned, which greatly improves the automatic reading accuracy of the instrument.

Description

technical field [0001] The invention relates to the technical field of automatic monitoring systems for instruments and meters, in particular to an automatic reading method for multi-dial pointer instruments. Background technique [0002] Due to the characteristics of simple structure, convenient use and easy maintenance, pointer instruments are widely used in various industrial fields. However, pointer instruments need to be manually read on-site. However, manual reading has a large workload, low efficiency, high error, and a series of uncertainties such as the reading accuracy is vulnerable to the working environment, which can no longer meet the needs of modern industrial production. Aiming at the problem of automatic reading of pointer instruments, researchers at home and abroad have proposed a series of algorithms including central projection method, subtraction method, template feature method, least square method, etc., among which central projection method, subtractio...

AI Technical Summary

Problems solved by technology

However, pointer instruments require manual on-site value reading, but manual reading has a large workload, low efficiency, high error, and reading accuracy is susceptible to a series of uncertain factors such as the working environment, which can no longer meet the needs of modern industrial production.

Aiming at the problem of automatic reading of pointer instruments, researchers at home and abroad have proposed a series of algorithms including central projection method, subtraction method, template feature method, least square method, etc., among which central projection method, subtraction method, etc. The method and the template feature method are greatly affected by the image shooting environment, and their robustness is poor, while the least square method has a large amount of calculation and high error, and cannot be applied to the occasion of real-time reading. In addition, the above methods can only be used for pointer instruments. Reading on a single dial requires a lot of image acquisition and storage during the reading process, which is not only inefficient, but also requires a lot of manpower and material resources

Images