Real-time video analytics on small autonomous drones poses several  difficult challenges at the intersection of wireless bandwidth,  processing capacity, energy consumption, result accuracy, and timeliness  of results. In response to these challenges, we describe four  strategies to build an adaptive computer vision pipeline for search  tasks in domains such as search-and-rescue, surveillance, and wildlife  conservation. Our experimental results show that a judicious combination  of drone-based processing and edge-based processing can save  substantial wireless bandwidth and thus improve scalability, without  compromising result accuracy or result latency.

Wang, J., Feng, Z.,  Chen, Z., George, S., Bala, M., Pillai, P., Yang, S., Satyanarayanan, M.
Proceedings of the Third IEEE/ACM Symposium on Edge Computing (SEC 2018), Bellevue, WA, October 2018