Pipeline reconstruction from fisheye images

Abstract

Automatic inspection of pipelines has great potential to increase the efficiency and objectivity of pipeline condition assessment. 3-D pipeline reconstruction aims to reveal the deformation of the pipe surface caused by internal or external influences. We present a system which can reconstruct the inner surface of buried pipelines from multiple fisheye images captured inside the pipes. Whereas the pipelines are huge, a fatal defect can be as tiny as a fine crack. Therefore a reliable system demands both efficiency and accuracy. The repetitive patterns on the pipe surface and the poor illumination condition during photographing further increase the difficulty of the reconstruction. We combine several successful methods found in the literature as well as new methods proposed by ourselves. The proposed system can reconstruct pipe surface not only accurately but also quickly. Experiments have been carried out on real pipe images and show promising performance.