Automated Detection and Classification of Infiltration in Sewer Pipes

Abstract

Many researchers have reported on the poor status of sewer pipes in Canadian cities, revealing the presence of a number of defects that impact their performance. Inadequate inspection is considered one of the main causes behind the declining condition of this class of pipes. This could be attributed to the high cost of inspection and inadequate funds allocated to this purpose. The labor-intensive current inspection practice gives rise to such high costs. This paper presents an automated system “AUTO-DETECT” that detects and classifies defects in sewer pipes automatically. The paper expands on earlier developments made by the writers, provides an overview of the overall configuration of the system, and focuses primarily on automated detection and classification of infiltration in sewer pipes. Infiltration is considered to be a serious and common defect that contributes to an undesirable extra flow of about 40% in this class of pipes. The system utilizes pattern recognition, image analysis techniques, and artificial intelligence to perform its task. A case example is presented to demonstrate the use and capabilities of the developed system.