Deep Learning-Based Industrial Meter Reading Recognition for Chiller Plants
In the AR inspection process of the chiller plant,to achieve automatic recognition of industrial meter readings and improve inspection efficiency,this paper proposes a chiller plant industrial meter reading recognition algorithm based on deep learning.First,images of the meters are captured using AR glasses.Then,a target detection algorithm is used to identify the meters in the images.The detected meters are cropped and input into an image segmentation algorithm,which segments the pointers and scales on the dial.Finally,the arc-shaped scales are converted into straight lines using polar coordinate transformation,and the final reading is determined based on the pointer position and the measurement range.This method has been applied in practical AR inspections of chiller plants with good results.
AR inspectiondeep learningmeter recognitionchiller plant
NETL
NSTL
万方数据
