Automatic identification method suitable for narrow scale pointer type instrument of transformer station

Abstract

The invention discloses an automatic identification method suitable for a narrow scale pointer type instrument of a transformer station. By detecting the ellipse circumference where scale straight lines are arranged, the scale straight lines pass a certain central center and based on the principle that profile segments corresponding to the scales appear in pairs, automatic number reading is achieved. In this way, dependency on manual calibration in an automatic identification process of a pointer type instrument can be reduced; intelligence of the identification algorithm is improved; the method can be applied to execution of task inspection of an inspection robot of the transformer station; and the method can be promoted to automatic identifications of other pointer type instrument like water meters, automobile oil level gauges and the like.

Description

technical field [0001] The invention belongs to the field of image pattern recognition, and in particular relates to an automatic recognition method suitable for narrow-scale pointer instruments in substations. Background technique [0002] The workflow of substation inspection robots to perform inspection tasks mainly includes: [0003] 1. Use the path planning algorithm to travel to the inspection point, and the gimbal drives the detection equipment to align with the monitoring instrument to be detected and capture the image of the instrument; [0004] 2. Use the meter recognition algorithm to analyze the meter image, obtain the meter readings, and then know the operating status of the equipment. [0005] In the whole inspection process, the meter recognition algorithm plays a decisive role, and its performance directly affects the quality of the inspection robot to complete the inspection task. Therefore, the design of the intelligent meter recognition algorithm is of gr...

AI Technical Summary

Problems solved by technology

[0007] For the automatic identification of pointer instruments in substations, the existing research mainly focuses on image preprocessing and pointer positioning. The identification of the dial is mainly through manual calibration on the template in advance. When the inspection robot performs the inspection task, the template is calibrated The parameters are matched to the inspection image, which has a great dependence on manual calibration. When the inspection robot fails to travel to the inspection point accurately, there will be deviations in the observation angle. The affine stretching degree of the instrument image to be observed There will be certain changes, and there may be some differences between the pre-manually calibrated instrument dial template parameters and the current actual instrument dial scale parameters. Therefore, when performing automatic readings, there will be certain errors in the reading results, making the obtained reading results inaccurate.

At the same time, because instruments with different ranges use different templates, it also makes the adaptability poor

Images