Purdue drone chases a person through forest without GPS or maps
Purdue's IDEAS Lab posted a 44-second video of a quadcopter following a person through woods with no GPS, map, or cloud computing. Navigation and occlusion prediction run on the drone's onboard computer. The footage overlays open-source SUPER and FAST-LIO components from HKU's MaRS Lab.
- Drone weaves between trees and keeps a person in frame without GPS or a forest map
- All processing runs onboard, avoiding cloud latency
- Occlusion prediction guesses where the target will reappear behind trunks
- Video shows open-source SUPER and FAST-LIO projects from HKU's MaRS Lab
Read next
Hardware