Font Size: a A A

Research On The Technology Of The Underground Pipeline Three-dimensional Detector Based On ARM Embedded

Posted on:2019-11-05Degree:MasterType:Thesis
Country:ChinaCandidate:K WangFull Text:PDF
GTID:2432330551961602Subject:Optical Engineering
Abstract/Summary:PDF Full Text Request
The pipeline three-dimensional detector detects underground pipeline,obtains its accurate spatial location distribution,provides service and guidance for urban infrastructure construction.In this paper,an intrusive pipeline three-dimensional detector is designed,and which is divided into pipeline robot device,pipeline position detection system and upper computer data processing system.The pipeline robot device carries pipeline position detection system and measures position information of the pipe axis.Upper computer data processing system to extract data to complete the reconstruction of underground pipelines.Pipeline position detection system is the core,And it uses single chip or single-core ARM processors,has shortcomings such as single thread,slow response,unable to process data in real time.Based on PSD and FBG detection method to measure pipeline deflection and curvature,which has the disadvantages of nonlinear,large error and low sensitivity.So,this paper designs a new pipeline position detection system,a quad-core ARM microprocessor,equipped with Linux operating system,multi-threaded,parallel computing,real-time filtering of acquisition signal.A new method of pipeline deflection and curvature detection is proposed,mainly using the camera to capture the laser spot,embedded image processing positioning laser spot position,to solve the pipeline axis curvature information.On the ARM embedded platform,the paper completed the three-axis accelerometer I2C driver,and accelerometer output signal is stable and reliable by offset correction and Kalman-filter processing;The laser spot collection and processing module,the first by Zhang's calibration method of camera calibration,and then gray transformation,adaptive median filteingr,threshold segmentation based on the most interclass variance,and finally the ellipses fitting algorithm to position the laser Spot coordinates.By simulating the underground pipeline experiment,the projection distribution of the pipeline in the horizontal plane and the vertical plane respectively is obtained.The mean absolute error of horizontal offset is 0.02992m and the average relative error is 6.7%.The mean absolute error vertical depth is 0.02984 and the average relative error is 5.1%.
Keywords/Search Tags:ARM embedded, underground pipeline detection, three-dimensional reconstruction, laser spot, image processing
PDF Full Text Request
Related items