Research Of The Micro Pipeline Robot Based On Machine Vision

In recent years,as the rapid development of industry and agriculture,the demand of pipeline is increasing year by year for it is widely used as an important way to transport water,oil,natural gas and other resources.However,due to the harsh environment in which pipeline usually installed for years,especially the influence of geological disasters,chemical corrosion and other external factors,it may lead to internal damage,leakage or even cause explosions and more serious incidents,which is a seriously potential danger to people's lives.To ensure that the pipeline will offer convenience to people's living and producing in the premise of safety and reliability,countries around the world constitute strict laws and regulations to force the corporate conduct pipeline inspection.Therefore,pipeline inspection robot was invented to inspect in the pipeline,replacing the inspector,which prevents injuries and improve the inspection effectiveness.In this paper,a spiral driven pipeline robot with remote monitoring function was independently designed.First of all,as for the kinematic mechanism design,compared with the advantages and disadvantages of current pipeline robot,combining with the physical condition of the inspected pipeline,the spiral driven mechanism with spring compression adjustment was designed and the test shows that it can stably walk in pipeline of various angles;as for the hardware system design,according to the fundamental requirement,the robot designed in this paper takes the PIC micro controller as a core controller,supplemented by hall sensor,ultrasonic sensor,dual-axis tilt sensor and other sensors and equipped with CCD network camera in front,it realizes the inspect of both the internal pipeline and the performance of robot.Secondly,based on the software development platform of Visual Studio 2010 and C#programming language,combining the hardware system,a software system for host computer at control center was designed to communicate with pipeline robot through wireless-fidelity.Software system can remote control the motion of robot,real-time display the specific information of pipeline and images of internal pipeline;as the expansion of the inspection,the design of six axis mechanical arm control system employs robot operating system as a platform,and C++ as programming language.Thirdly,based on the defect images of the internal pipeline,this paper designs image processing algorithm dealing with the corrosion image within pipeline.Because of the feature of camera imaging,cylinder unwrapping algorithm is used for collected internal pipeline images,and LAB color space is adopted to avoid the influence on imaging of device channel,after that,block + Otsu + K-means algorithm is used for image segmentation.The research shows that this algorithm is much better than others when processing the defect image of pipeline on the segmentation quality,and much quicker on computing speed.