OGInfra: Geolocating Oil & Gas Infrastructure using Remote Sensing based Active Fire Data
This addresses the need for efficient monitoring of oil and gas infrastructure for environmental or regulatory purposes, though it appears incremental as it applies deep learning to a specific domain.
The paper tackled the problem of automatically geolocating oil and gas infrastructure using remote sensing data, achieving a top accuracy of 90.68% with ResNet101.
Remote sensing has become a crucial part of our daily lives, whether it be from triangulating our location using GPS or providing us with a weather forecast. It has multiple applications in domains such as military, socio-economical, commercial, and even in supporting humanitarian efforts. This work proposes a novel technique for the automated geo-location of Oil & Gas infrastructure with the use of Active Fire Data from the NASA FIRMS data repository & Deep Learning techniques; achieving a top accuracy of 90.68% with the use of ResNet101.