Research On Deformation Detection And 3D Reconstruction Of Pipeline Based On Active Panoramic Vision

Traditional ways of transportation include railway,highway,water transport and air transport,and the pipeline transportation as another type of shipping,especially for the transportation of chemical products,easily prone to defects such as corrosions.In view of the above reasons,it is necessary to develop an automatic and intelligent health monitoring technique to maintain the pipeline on a regular basis.Therefore,a kind of 3D omni-directional vision detection method and the corresponding sensor which is suitable for the pipeline inspection is presented in this thesis.The main contents and achievements of this thesis include:1.According to the characteristics of small internal space and long laying distance of pipelines,a kind of pipeline crawling robot which can walk freely and inspect quickly in the pipeline is adopted in the thesis.2.In order to obtain the panoramic image inside the pipeline,a kind of O mni-Directional Vision Sensor(ODVS)which can obtain a larger vertical field of view even in a small space is presented in this thesis.3.In order to measure the depth information inside the pipeline accurately,the ODVS is combined with a panoramic laser generator to form an Active Stereo O mni-Directional Vision Sensor(ASODVS),which can achieve a 3D measurement and reconstruction of the pipeline according to the principle of structure light vision.4.In order to detect the morphology defects of the pipeline,the image data include texture information and depth information is required,thus the ODVS is combined with the ASODVS to form an active binocular omni-directional vision sensor,meanwhile,the panoramic laser generator is replaced by a panoramic infrared generator and a light filter is setted in the middle of the ODVS and the ASODVS to separates the infrared light and the visible light based on the characteristics of the wavelength.Finally,the device can shoot the panoramic texture image and panoramic infrared scanning image simultaneously.5.In order to overcome the problem of the measurement error caused by the roll of the crawling device,we use the panoramic texture images to estimate the motion poses of the panoramic camera in real-time to correct the coordinate system of the 3D point cloud data.6.In order to make a more accurate assessment of the degree of the damage,we calculate the quantification information of the defect,including the location,surface area and volume.7.In order to detect the deformation o f the pipeline in real-time,the deformation analysis of the point cloud data is carried out,including the ca lculation the cross-sectional area and the deformation rate of the pipe.In this thesis,the active binocular omni-directional vision theory is used to detect the structural and functional defects of the pipeline.The experimental results show that the detection device designed in this thesis can quickly obtain the texture information and geometry information inside the pipeline,the point cloud data is able to quantify the defect accurately,detect the deformation in real-time and reconstruct the pipeline precisely.