A key point detection method of instrument pointer

Abstract

The invention discloses a method for detecting key points of an instrument pointer, which belongs to the field of image recognition. The invention discloses a method for detecting key points of an instrument pointer, comprising the following steps: Step S00: data collection of substation pointer-type instrument pointer dials, and robot inspection The substation takes a picture of the pointer meter; step S10: data annotation and data augmentation, including pointer table rectangle frame, pointer key points, and labeling of dial scale key points; step S20: constructing a deep network including a target detection network and a key point detection network ; Step S30: the pointer meter picture is input into the target detection network after size reconstruction; Step S40: the pointer meter picture after the size reconstruction is rotated, scrambled and enhanced, and then intercepted to generate a full-angle data set and input to the key point detection network; Step S50: Calculate the scale estimation value of the pointer meter. It gets rid of the instability caused by noise and light distortion in traditional algorithms, and has robustness and ease of use in use.

Description

technical field [0001] The invention relates to the field of image recognition, in particular to a method for detecting key points of instrument pointers. Background technique [0002] The current detection and recognition methods of instrument scale pointer include traditional image recognition technology methods and recognition methods based on deep learning. Traditional image recognition methods are divided into template matching method, maximum connected domain method, silhouette method, least squares fitting method, line drawing method and region growing method, etc. The method based on deep learning has the advantages of robustness, accuracy, and simplicity, and has become a research hotspot in recent years. Usually, the deep learning framework is used to locate the rectangular frame of the pointer scale area of ​​the dial, and then the traditional detection scale reading and pointer angle are used in the rectangular frame. The latest research is based on deep learni...

AI Technical Summary

Problems solved by technology

[0003] Due to the influence of the production environment light, there are differences in the light angle and intensity at different times. In the traditional algorithm, the recognition is unstable due to the influence of noise, light distortion, etc.

Images