Sensing for Feature Identification in Sewers

Multi-disciplinary research programmes have been established at Sheffied Hallam University to investigate specific areas of the Information Technology into the Construction industry theme. Interest is currently centred upon the development of suitably robust knowledge-based and related engineering systems for application in non-man-entry (NME) sewer renovation. Initial efforts have been directed towards provision of an appropriate vision sensing facility. A task dedicated computer vision system combining edge detection and grouping algorithms with a controllable light source is proposed. The goal of this system is to extract descriptions of lateral connections from images captured during the survey pass of an Insituform process. Initial work is focused on the problem of lateral detection.