Machine learning for the automated detection of deer in drone and camera trap imagery

To effectively manage the growing population of invasive deer in Australia, adequate monitoring techniques are essential. Traditional methods of detecting and monitoring deer such as scat surveys, spotlighting, or piloted aerial surveys can be expensive and time consuming. To overcome these issues, camera traps and remotely piloted aircraft systems (RPAS or drones) are increasingly being used to detect and monitor deer populations. This thesis presents a new method for assessing the imagery provided by RPAS and camera traps using Machine Learning, reducing the time and cost of assessing deer populations, providing the opportunity for more efficient management