Overview
Ring-billed gulls (Larus Delawareans) and Herring gulls (Larus Argentatus) are
ranked as some of the riskiest wildlife species for civil and military
aviation. The August 2019 Ural Airlines Airbus 321 passenger flight from
Moscow, Russia that crash landed in a cornfield due to a collision with a gull
flock shortly after takeoff highlights the need for gull management techniques
to improve aviation safety. Fortunately, all 226 passengers and crew survived
the crash; however, the risk of gull flock collisions still exist. Finding
ways to safely disperse gull flocks near airports is an important step in
reducing these risks.
This goal of this study is to develop computational tools to study the escape
responses of both Ring-billed and Herring gulls when being approached by three
types of sUAS platforms (remote-controlled, quadcopter, and an ornithopter).
Previous studies on the escape responses of birds such as Turkey Vultures
(Cathartes Aura) who typically travel in small numbers (usually 1-6) have
utilized unautomated video to collect data of flock flight-initiation distance
and reaction time. Because gulls travel in much larger numbers (often
hundreds), computer visualization techniques will be used to identify and track
individual gull escape reactions. Multiple object tracking (MOT) will allow
for automated collection of flight-initiation distance, reaction time, takeoff
speed, and angle of escape for each of the gulls in the video.