Font Size: a A A

Research On Oil Leakage Detection System Of Hydraulic Components Based On Machine Vision

Posted on:2019-03-26Degree:MasterType:Thesis
Country:ChinaCandidate:X K SunFull Text:PDF
GTID:2382330596950124Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
In the hydraulic system,oil leakage is a typical repeated problems.Oil leakage will not only cause economic losses and environmental pollution,but also may cause fire and explosion.How to detect oil leakage accurately and automatically is a complicated and challenging task.At present,most domestic enterprises are still using manual testing methods,which can not detect oil leakage timely,fast and accurately.So using machine vision to detect oil leakage is an inevitable trend in the future.From the perspective of machine vision,the possibility of constructing a set of hydraulic components oil leakage detection system which is based on machine vision is discussed in this thesis.The main contents of the research are as follows:Firstly,After spatial transformation model was researched and analyzed,an image registration algorithm based on SURF features and CPD point sets is proposed in the thesis,which can effectively solve the problem that the hydraulic components are easily moved and rotated in the image acquisition process and improve the accuracy of the subsequent image difference method.Afterwards,The gap images taken under auxiliary light source usually have serious loss of visibility and contrast,which make it difficult to detect oil leakage.To solve the problem,traditional image enhancement algorithms were studied,an enhancement algorithm of gap image is proposed based on improved dark channel prior,The experimental results demonstrate that the proposed approach can effectively improve the perceptual quality of gap image.According to the feature that the oil leakage is a dynamic process,an image segmentation algorithm based on the image difference method is proposed,which can effectively detect the changed areas in the image.Then,the feature of these regions are extracted,and finally the method of support vector machine is used to identify oil leakage.Finally,According to the technical requirements of oil leakage detection system for hydraulic components,the overall design of oil leakage detection system was proposed and the hardware structure and software of the system were built.The experimental study on the oil leakage of hydraulic components was carried out.The stability and accuracy of the system are tested.The experimental results show that the system can detect the oil leakage of hydraulic components quickly and accurately.
Keywords/Search Tags:Machine Vision, Image Enhancement, Image Registration, Feature Extraction, Oil leakage
PDF Full Text Request
Related items