Motivation

Understanding the dynamics between a shepherd, sheepdog and a flock of sheep is crucial for efficient and low-stress livestock management. Traditional methods of studying these behaviours are often qualitative and labour intensive. There is an opportunity to apply modern sensing and analysis techniques to gain insights into herding behaviours. This project aims to use aerial drone footage to analyse the complex interactions within a herding environment automatically.

The core objective is to develop a system that can detect and track individual agents, sheep, dogs and people, from video data. By analysing their positions and movements over time, the project will compute key flocking metrics, such as the “flight radius” (the distance at which sheep begin to move away from a dog or person) and the “influence radius” (the area a dog can effectively control). These quantitative metrics will provide valuable data for optimising herding strategies, improving animal welfare and could serve as a foundational dataset for training autonomous robotic shepherds.

Required Skills

• Programming skills in Python.
• Mathematics, 2D geometry for calculating distances and positions.
• Basic knowledge of machine learning and computer vision principles (e.g., object detection, tracking).
• Experience handling and processing video data and/or sensor data (such as GPS) would be beneficial.
• Familiarity with computer vision libraries such as OpenCV and deep learning frameworks like PyTorch or TensorFlow is advantageous.
• An analytical and problem solving mindset.

Skills to Be Gained

This project provides an opportunity to apply technology to the field of animal behaviour and precision agriculture. You will gain hands-on experience with a project involving computer vision, data processing, deep learning model implementation and analysis. You will develop skills in object detection and tracking algorithms, as well as techniques for fusing video data with other sensor inputs, such as GPS.

This project is suitable as a final year project for students at Lincoln studying Computer Science, Games or Robotics, or as an internship in robotics research. If you are interested, fill out our Expression of Interest Form, choosing Dr Jonathan Cox (jcox@lincoln.ac.uk) and Dr Rajitha de Silva (ODeSilva@lincoln.ac.uk) as the researchers to supervise the project.