From bees to Drones: Exploring bio-inspired machine vision applications for precision agriculture
September 28 @ 6:00 pm - 7:30 pmFree
Bees are used as vectors for pollination and transport of agricultural chemicals in outdoor agriculture and greenhouses. However, in certain situations the use of natural pollinators is problematic, namely: hybrid production, cultivation of pharmaceutical drugs, chemical transport greater than the tolerance of the vector species, or atmospheres poor in oxygen.
Small unmanned aerial vehicles (UAVs) could serve as an alternate pollination vector in such situations and perform other functions (i.e. yield estimation, early pest detection, soil moisture estimation, etc.). A step towards the deployment of such a tool is to bring the ability to locate, classify, and analyze flowers aboard a drone. Current machine vision approaches in agriculture focus on the visual and infrared spectrums. This project takes a novel approach by analyzing images from the UV spectrum; better mimicking the visual system of natural pollinators for locating and distinguishing flowers. During evolution, flowers have developed reflective and UV absorbing compounds to better distinguish their reproductive structures from conspecifics and various natural environments. This allows potential pollinators to locate the species and the floral part to be visited at a relatively large distance while traveling at an efficient speed.
This talk will introduce the concept of precision agriculture, the current state of robotic “bees” in development, and the logistics/benefits of doing image analysis in the UV spectrum. Preliminary results from this study will also be presented as well as plans for upcoming field tests.
Megan Heath is a Master’s candidate in environmental engineering at École de technologie supérieure. Her work focuses on finding modern, technological solutions to environmental problems. She has recently been accepted into the NSERC- funded UTILI CREATE program for training on UAS systems. She derives inspiration for her work from her passion for the outdoors and vintage Sci-fi.
Lab page link: https://initrobots.ca/recherches/voir-comme-une-abeille
LinkedIn profile: https://www.linkedin.com/in/megan-heath-a7b24494/