Geospatial Research Group

The Geospatial Lab performs research in a variety of remote sensing topics, from the development of novel machine-learned models, to multimodal data fusion, and beyond. Through the utilization of high-resolution multispectral remote-sensed imagery, a multimodal space-based vehicle detection capability has been developed. Through middle fusion, detection capabilities are enhanced through data sharing. The Lab is actively developing multimodal ML-enhanced UAS systems, capable of executing ML models in real-time. The object detection model will utilize NIR hyperspectral sensing, LWIR imagery, and standard RGB imagery in a multimodal YOLO-based system, operational on a COTS UAS. We have developed remote sensing pipelines to detect large-scale perturbations to critical infrastructure by utilization of synthetic aperture radar imaging.