Visual biometrics

BVI research led by Tilo Burghardt aims at providing non-invasive approaches to assist with problems in field biology, for example, to better understand and conserve endangered species. Specifically, the team has developed approaches that facilitate remote monitoring and identification of individual animals in large populations using computer vision and biometric techniques.

Building on experiences from proof-of-concept work on visual lion tracking, early work supported by The Leverhulme Trust demonstrated that subpopulations of African penguins could be recorded in their natural habitat enabling individuals to be identified visually by their coat markings. More recently, this concept of ‘animal biometrics’ has been used to help monitor other endangered species such as great apes and sharks. In the latter case, the biometric is based on the characteristics of the shark’s dorsal fin. If high-quality imaging is possible and animals carry sufficiently complex markings, this approach works very robustly in dealing with different scales, viewpoints and even partial occlusions.

The Save Our Seas Foundation has recently started building on this research by integrating the developed white shark software with the global online platform Wildbook in order to allow researchers to submit shark fin photographs and match their sightings against large population databases. Animal biometric systems for great ape research are currently being developed further with German collaborators including the Max Planck Institute for Evolutionary Anthropology in Leipzig with the goal of large scale application to camera trap footage.

3.4 Visual biometrics secondary
Edit this page