Exploring the Role of Deep Learning and Computer Vision for Monitoring On-Farm Biodiversity.
Dr Roslyn Henry, University of Aberdeen; Prof Georgios Leontidis, University of Aberdeen; Dr Mamatha Thota, University of Lincoln; Prof Jonathan Hillier & Dr Milad Toolabi, Trinity AgTech
Halting the accelerating loss of biodiversity is one of the greatest pressures of our time. The Global Biodiversity Framework (GBF) has set out ambitious global targets to address the biodiversity crisis by 2030. Agriculture dominates 44% of the world’s habitable land and therefore has an important role to play in meeting biodiversity conservation targets. As such we need tools for estimating and monitoring biodiversity in agricultural settings to increase the sustainability of the agri-food sector.
Measuring biodiversity on the ground is labour intensive and difficult to deploy at large scales. Alternatively, earth observation data (EO) from satellites can provide image data for use within biodiversity monitoring and is an ideal data source as it is available at large scales, low cost and updated regularly. However, techniques for modelling and monitoring biodiversity change from remotely sensed data remain under-developed. This PhD, based at the University of Aberdeen, in collaboration with The University of Lincoln and Trinity AgTech, will address this knowledge gap and conceptualise new artificial intelligence methods to develop biodiversity monitoring tools for agricultural settings. A comprehensive training programme will be provided comprising of both specialist scientific training and generic transferable and professional skills. Advanced training in AI techniques will be tailored to address the needs of the student and the student will have the opportunity to attend courses to fill training gaps. The student will work closely with Trinity AgTech and their natural capital navigator, Sandy. As such the student will have the opportunity to learn about the use of biodiversity monitoring in industry and potentially contribute to advancing the representation of biodiversity in natural capital markets.