Assessing the Feasibility of Identifying Wildlife from A Drone’s Perspective

Rozina Myoya , Lesego Matojane

Partner: para-zoology

Year: 2023

Abstract: This project aims to identify the unique features of crocodiles from drone-derived imagery and analyse the potential of using this technology for monitoring and conserving crocodile populations. The project will use a combination of computer vision and machine learning techniques to conduct animal detection and individual re-identification. The focus will be on the detection of crocodiles within the images, and thereafter the recognition of individual crocodiles through the identification of their unique morphological features. This information will be used to create a database of unique features for each crocodile that can be used for population monitoring and re-identification in the wild. The project will also explore the limitations and challenges associated with using drone-derived imagery for crocodile monitoring such as the resolution of the imagery and the difficulties in distinguishing between crocodiles and other similar-looking animals or objects. Ultimately, the results of this project will provide valuable insights into the potential of using drone technology for monitoring and conserving crocodile populations, and contribute to the development of effective and sustainable conservation strategies.

Presentation Video