ABSTRACT
Biodiversity conservation and environmental monitoring are pressing challenges in the context of climate change and the accelerated loss of species. In this scenario, computer vision and deep learning technologies offer new opportunities to automate tasks such as species identification, counting, and tracking, as well as the analysis of animal behavior across different environments.
The main objective of this PhD research is to design and validate computer vision-based methodologies and tools that improve biodiversity and environmental monitoring. The specific goals include: (i) real-time species counting and habitat monitoring, (ii) behavioral analysis of abandoned domestic animals, (iii) development of scalable ecological surveillance systems using edge-cloud architectures, and (iv) a critical evaluation of the use of large language models (LLMs) for environmental monitoring and misinformation detection.
The research combines deep learning, object detection and tracking, multithreaded video processing, and automated environmental data analysis. In addition, reproducible methodological frameworks have been developed to leverage the capabilities of distributed computing in edge-cloud environments.
The results demonstrate the feasibility of designing effective and adaptable solutions for monitoring in both natural and urban settings. The thesis presents significant advances in automated species counting, tools for automated animal behavior analysis, and a critical assessment of the strengths and limitations of language models in environmental applications.
This work contributes to advancing research in computer vision for sustainability and provides useful tools for conservation projects, wildlife management, and real-time environmental monitoring.
PhD Tutor:
PhD Advisors:
PHD CANDIDATE
Oluwakemi A. Akinwehinmi
is rounding off her PhD in Information Technology and Engineering from Universitat de Lleida, Spain (2022–2025). She holds a Master of Science in Computer Science (2021) and a Bachelor of Science in Computer Science (2016) from the University of Ibadan, first and best University in Nigeria. Her doctoral research explores the application of computer vision and artificial intelligence to biodiversity and ecological monitoring, with an emphasis on deep learning models for real-time detection, environmental surveillance, and data-driven decision-making.
She was Research Assistant at the International Center for Numerical Methods in Engineering (CIMNE), Spain during her doctoral studies, where she contributes to national and international research projects funded by the Government of Catalonia and collaborative institutions. Her work includes the development and evaluation of machine learning models for habitat monitoring, abandoned animal detection, and the integration of heterogeneous environmental data streams. In 2023, her research team received the Joan Roget Knowledge Transfer Award for excellence in collaborative R&D. In addition to her academic career, she has industry experience in IT infrastructure management, cloud services, and data systems. Her research interests include computer vision, machine learning, deep learning, biodiversity monitoring, and AI-powered data integration for real-world applications.