Transformer-substation wide-scale pointer type instrument automation identification method

Abstract

The invention relates to a transformer-substation wide-scale pointer type instrument automation identification method. The digital scale form of a wide-scale pointer type instrument is a rectangle, and a digital scale interval is 1. The method comprises the following steps of firstly, reading an instrument image and preprocessing the image; then, identifying an instrument scale, which includes theautomatic positioning of each rectangular instrument scale in the image, carrying out scale radial angle ordering and realizing the establishment of an instrument scale coordinate system; and then, using an LSD straight line detection algorithm and a scale distribution ellipse circumference constraint to carry out pointer identification and positioning; and finally, realizing the automatic interpretation of an instrument and reading and displaying a number. Dependence on people in an identification process is subtracted. During the identification process, extra instrument scale parameter input is not needed. Under the condition of guaranteeing identification precision, the rapid reading of the wide-scale pointer type instrument is realized. The method can be used by a transformer substation inspection robot to execute an inspection task and has good intelligent performance.

Description

technical field [0001] The invention relates to the field of image pattern recognition, in particular to a method for automatic recognition of wide-scale pointer meters in substations. technical background [0002] Although digital instruments are more and more widely used, pointer instruments are still widely used in substations with harsh environments because of their strong anti-interference ability and high reliability of display recognition. Pointer instruments usually do not have the function of data line interface, and cannot automatically identify and return the readings in the meter. Therefore, the data recording of pointer instruments is mainly based on close-up observation by human eyes and manual recording. Repeated mechanical movements can easily cause human visual fatigue, resulting in errors in identifying and recording data. With the advancement of automation technology in substations, it has become a trend to use inspection robots to collect pictures of poi...

AI Technical Summary

Problems solved by technology

For the template matching method, it is necessary to manually calibrate the meter on the picture containing the meter first, and then match the calibrated meter on the picture to be recognized. This not only limits the types of meters, but also relies heavily on manual calibration, which is not conducive to The Development of Substation Automation

For the Hough transform, the traditional Hough transform algorithm has a long operation time and a large storage space, which makes the recognition process inefficient.

[0004] Due to the influence of the light and the shooting angle of the inspection robot, when the inspection robot is used to collect the pictures of the pointer meters in the substation, the meters in the pictures are usually stretched, which will affect the recognition and accurate reading of the dashboard

Images