A Fast and Low-cost Method to Automate Detecting, Locating, and Mapping Internal Gas Pipeline Corrosion using Pig-mounted Thermal and Stereo Cameras

Main Objective

To test and evaluate performance of the proposed automatic multispectrum imaging fusion method in detecting and locating internal gas pipeline corrosion through lab and field trails.

Public Abstract

An automated multi-channel imaging fusion approach is proposed to create an internal inspection device, which can conduct fast, lowcost, accurate, and automated detecting, locating and mapping internal gas pipeline corrosion and presence of water. This novel approach is expected not only to significantly reduce the discovery time and cost of inspections, but also to enable corrosion formation monitoring by comparing chronological corrosion inspection records due to the precise mapping capability.
An inline pig-like device mounted with both thermal and stereo cameras will be designed. The back-end mounted thermal camera will record continuous thermal images. The front-end mounted stereo cameras will record visual and texture surface images. These thermal and visual data will feed into a developed image stitching, mapping and analysis module to automatically generate defect thermal characterization and 3D profiling of the entire internal pipeline. Corrosion spots and leaking areas will be detected, located, mapped, and evaluated automatically in the post processing stage.
The feasibility of the proposed approach is demonstrated with preliminary results. This in-service inspection device has five notable advantages: 1) easy to use in the field; 2) fast travel over long distances; 3) no service interruptions; 4) automated corrosion detecting, locating, and characterization; 5) corrosion formation monitored by time-series analysis by comparing the chronological inspection images.
Successful implementation of this research could save the gas industry billions of dollars in the maintenance of transmission and distribution lines. The lives saved, and environmental benefit are difficult to qualify.