Using new imaging methods to investigate rare mammals in historical museum collections
Lead supervisors: Dr Jennifer Hoyal Cuthill
Location: School of Life Sciences, University of Essex
Duration: 8 weeks
Suitable undergraduate degrees: Biology, geology, computer science or related subject
Large public databases hosting three-dimensional scans of mammal skulls, together with new methods of machine learning, developed by the lead supervisor (Hoyal Cuthill et al, 2019, Science Advances; 2020 Nature) promise new insights into the form, function and evolution of mammal morphology. However, some of the mammal species of greatest evolutionary and conservation interest are absent or under-represented in online digital collections, yet are available for study in the UK’s large historical museum collections. Examples include the sirenians (manatees and dugongs) a morphologically unusual group that includes both vulnerable and recently extinct species, represented by only a single digital specimen on the public DigiMorph database. The proposed project will help fill such data gaps by trialling a new photographic method for generating data comparable to CT still images, using RTI (reflectance transformation imaging), which uses post-processing of multiple photographs to generated images in which the lighting and colour can be computer controlled. After photographic training by the project supervisor at the University of Essex, the placement student will visit the collections of the University Museum of Zoology, Cambridge to systematically photograph museum specimens of mammal skulls held in the collections and make digital records of accompanying specimen label data. They will also be trained to process images, at the University of Essex, using cutting-edge computer analysis techniques. The placement project will contribute to a major research project in development by the project supervisor, using machine learning to understand the form, function and evolution of mammal morphology.