Tracking Honey Bees with Computer Vision through the use of Fiducial Markers and Bayesian Inference

• Grant number: 23/05186-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): August 01, 2023
• Effective date (End): January 31, 2024
• Field of knowledge: Engineering - Electrical Engineering
• Principal Investigator: Carlos Dias Maciel
• Grantee: Maurício Garcia di Mase
• Host Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated research grant: 14/50851-0 - INCT 2014: National Institute of Science and Technology for Cooperative Autonomous Systems Applied in Security and Environment, AP.TEM

Abstract

Honey bees are insects that, despite being simple at an individual level, show complex social behavior and communicate with each other and with the queen bee through sound signals, vibration in the honeycombs and even through behavioral signs (e.g. waggle dance) (Seeley, 2010). Developments in the area of precision apiculture and in biological research about the susceptibility of these insects to pesticides broadly used in the brazilian agriculture, such as imidacloprid, has highlighted the relevance of computational algorithms capable of monitoring bees or tracking them through computer vision (Eren et al., 1997; Uthoff et al., 2023; Qandour et al., 2014; Ramsey et al., 2020; Campbell et al., 2008; Ngo et al., 2019; Boenisch et al., 2018; Wario et al., 2015; Bozek et al., 2021; Crall et al., 2015; Júnior et al., 2019). In this research project, the author proposes to implement a monitoring algorithm that relies on the use of a pre-processing based in the identification of individualized fiducial markers placed on the honey bees and a processing based on bayesian inference in order to track each marked bee individually. This approach based on baysian probabilistic programming has precedents in various other solutions in data analysis and is in direct continuity with the line of research of the signal processing laboratory (LPS-SEL-EESC) (Villanueva, 2012; Endo et al., 2014; Santos, 2017; Nakashima et al., 2017; Gross et al., 2019, 2018; Oliveira Jr et al., 2022), which takes part in the INCT-SAC, to which this project is intended to be bound to. Once tested and fully functioning, the tool may be used by researchers of the Federal University of Viçosa, which will already provide recordings of honeybees to test the algorithm.