A new system developed by University of California, Santa Barbara (UC Santa Barbara) researchers uses hyperspectral imaging and machine learning for more accurate detection of methane emissions.
The researchers used data from hyperspectral cameras in the vicinity of the Four Corners region of the American Southwest, where methane emissions are high. Hyperspectral imaging can capture spectral “fingerprints” that correspond to methane's 2,200-2,400 nm spectral band.
Said UC Santa Barbara's Satish Kumar, "We used a deep learning model to train the computer to learn the shape that a methane gas leak takes as it is released and spreads," which helped to pinpoint the location from which methane was being emitted, and to differentiate between methane and other hydrocarbons in the same image.
From The Current (UC Santa Barbara)
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA
No entries found