Benjamin E Natarian
Download Full Text (604 KB)
Currently detecting a natural gas leak requires a boots on ground approach with handheld and vehicle mounted chemical sensors. The goal of this research is to create an automatic detection system using manned or unmanned aircraft to survey a large area in a short time and narrow the leak location to a minimal search area for boots on ground isolation. The proposed system will utilize an IR sensor for data acquisition, and a custom computer vision based algorithm to detect gas leaks in the massive data stream with minimal human interaction. Specifically, the algorithm takes a combination of preprocessing, motion estimation, and machine learning to differentiate between gas leaks and normal background.
Graduate Research - Graduate
Vijayan K Asari
Primary Advisor's Department
Electrical and Computer Engineering
Stander Symposium poster
"Remote Sensing: Natural Gas Detection with Infrared (IR) Spectrum" (2017). Stander Symposium Posters. 901.