Automated inspection of sewers using ultrasonic sensors

摘要

Ultrasonic sensing is a relatively new method of sewer inspection. The most common sewer inspection used today is based on closed circuit television (CCTV) where a video camera is mounted on an inspection robot. Such a system provides images to an operator, who detects, classifies and rates the severity of defects against documented criteria. However, very little work has been done on a fully automated sewer inspection. Instead, most research has been directed at aiding the manual interpretation process to determine the defect characteristics. The paper deals with the initial findings on the use of an underwater ultrasonic inspection system. An artificial neural network is employed for detecting faults in the inner wall of the inspected sewer segment. (9 References).