Combined Person Classification with Airborne Optical Sectioning
This addresses the challenge of finding lost or injured persons in search-and-rescue operations, representing an incremental improvement over existing methods.
The paper tackled the problem of detecting persons under occluding forest canopy using drones, achieving improved classification rates by combining classifications from multiple Airborne Optical Sectioning images, with real-time processing at groundspeeds up to 10 m/s.
Fully autonomous drones have been demonstrated to find lost or injured persons under strongly occluding forest canopy. Airborne Optical Sectioning (AOS), a novel synthetic aperture imaging technique, together with deep-learning-based classification enables high detection rates under realistic search-and-rescue conditions. We demonstrate that false detections can be significantly suppressed and true detections boosted by combining classifications from multiple AOS rather than single integral images. This improves classification rates especially in the presence of occlusion. To make this possible, we modified the AOS imaging process to support large overlaps between subsequent integrals, enabling real-time and on-board scanning and processing of groundspeeds up to 10 m/s.