Font Size: a A A

Research On Method Of Underwater Pipeline Detection And Location Estimation Based On AUV

Posted on:2018-02-19Degree:MasterType:Thesis
Country:ChinaCandidate:Y H LiuFull Text:PDF
GTID:2310330542487176Subject:Ships and Marine engineering
Abstract/Summary:PDF Full Text Request
The ocean has a large amount of mineral resources,however the exploitation of marine resources needs professional equipment and technical support due to the special underwater environment.The Autonomous Underwater Vehicle(AUV)is an underwater operation equipment which can achieve the underwater operation autonomously.Adopting the optical vision system,an AUV can complete various tasks such as underwater environmental detection,target recognition.Nowadays,the AUV technology is a research hotspot around the world,and the underwater target detection is a key technology in the process of research.In this paper,the method of underwater pipeline detection and position estimation based on AUV has been studied.The main research contents are as followed:According to the principle and characteristics of underwater optical imaging,the image processing algorithm of underwater pipeline detection has been discussed from four aspects: filtering,enhancement,segmentation and edge detection.In consideration of gray level missing and big calculation of the traditional fuzzy enhancement algorithm,the traditional fuzzy enhancement algorithm has been improved from two aspects: membership function and operation speed.The methodology of edge feature filter method based on region saturation control has been presented,which can reduce the influence of interference pixels on the feature extraction.The underwater image processing flow has been produced by analyzing the characteristics of underwater image and the rapidity of the algorithm.On the basis of the traditional Hough transform pipeline detection method,a pipeline detection method based on linear feature clustering has been studied.Using the improved Hough transform linear feature of binarization image of pipeline can be extracted,the improved K-means clustering algorithm is adopted to cluster the linear feature,then limit the clusters according to the position of pipeline detected from previous image,finally position of the pipeline central axis is got.This method can effectively avoid the influence of the edge information extraction error on the accuracy of pipeline detection.Based on groups of pipeline sequence diagram,the off-line detection experiment is carried out,compared with the traditional pipeline detection method,the superiority of this method has been proved.Research on location estimation method of underwater pipeline based on AUV.On thebasis of the camera imaging model analysis,the imaging geometric model of the underwater pipeline detection and tracking system has been constructed,and the conversion relationship between the coordinate systems has been derived.The internal parameters of the camera has been obtained from the underwater camera calibration test.The position of pipeline relative to the AUV has been estimated according to the data of pool experiment,the effectiveness of pipeline position estimation method has been confirmed.
Keywords/Search Tags:AUV, pipeline detection, Hough transform, position estimation
PDF Full Text Request
Related items