Font Size: a A A

Design Of Pointer Type Of Water Meter Precision Testing System Based On Machine Vision

Posted on:2019-10-09Degree:MasterType:Thesis
Country:ChinaCandidate:D XingFull Text:PDF
GTID:2392330596994795Subject:Electrical engineering
Abstract/Summary:PDF Full Text Request
The measurement of water consumption with water meters is a matter of concern in daily life and industrial water use.At present,most of the residents and factories in our country still use mechanical pointer water meters,and the accuracy of the water meters needs to meet the requirements of the calibration regulations before they come into service.Nowadays,in the water meter manufacturing industry,the traditional precision testing method is still adopted by most enterprises by way of reading the flow point value of water meter manually.The accuracy of the testing data of the method is naturally low,and the manual operation cost is large,so it can not meet the task of detecting a large number of water meters in the enterprise.However,the improved detection methods such as electromagnetic flowmeter are still of the disadvantages of poor sensitivity.In this paper,an automatic water meter precision detection system based on machine vision technology is design to improve the accuracy of the detection results.The main investigated work of the paper is as follows:1.According to the needs of the work,the overall scheme of the system is designed.The accurate detection method of water meter is put forward,and the vision imaging scheme is studied.Then the Camera,lens and light source are carried out according to the vision imaging project.Mechanical model is constructed.Lastly the hardware circuit selection is contrived.2.The image processing algorithm is probed,including image pre processing,extraction of pointer region information,corner recognition and reading calculation.The corner recognition as for the original Harris corner detection algorithm mistakenly identified a number of false corners is researched.Therefore,an improved Harris corner detection algorithm is proposed by adding bilateral filtering and two non-local Maxima suppression based on confidence of comparison,which can eliminate false corners and screen the only accurate corner.3.The development of testing software is finished.The overall framework of the system detection software is examined.Complete the development of camera image acquisition tool and coordinate calibration of function are completed,and the software information management platform is built.4.The system scheme is tested in the industrial field.The results show that the design can meet the requirements of enterprise water meter precision.Compared with the results of manual precision testing,the deviation of data sample is reduced and the accuracy is improved.
Keywords/Search Tags:water meter accuracy detection, machine vision, Harris corner detection algorithm, Non-local Maxima suppression
PDF Full Text Request
Related items