An automated program to find animals and crop photographs for individual recognition

Author(s)Patrick Buehler, Bill Carroll, Ashish Bhatia, Vivek Gupta and Derek E. Lee
Year Published2019
JournalEcological Informatics
Page Numbers191-196
Size2.09 MB

Detailed data on individual animals are critical to ecological and evolutionary studies, but attaching identifying marks can alter individual fates and behavior leading to biases in parameter estimates and ethical issues. Individual-recognition software has been developed to assist in identifying many species from non-invasive photographic data. These programs utilize algorithms to find unique individual characteristics and compare images to a catalogue of known individuals. Currently, all applications for individual identification require manual processing to crop images so only the area of interest remains, or the area of interest must be manually delineated in each image. Thus, one of the main bottlenecks in processing data from photographic capture-recapture surveys is in cropping to an area of interest so that matching algorithms can identify the individual. Here, we describe the development and testing of an automated cropping program. The methods and techniques we describe are broadly applicable to any system where raw photos must be cropped down to a specific area of interest before pattern recognition software can be used for individual identification. We developed and tested the program for use with identification photos of wild giraffes.

Keywords: Animal detection, Capture-recapture, Computer vision, Histogram of oriented gradients, Image processing, Individual identification, Photo identification

Authors: Patrick Buehler, Bill Carroll, Ashish Bhatia, Vivek Gupta and Derek E. Lee

Journal: Ecological Informatics

Terms and Conditions: Any PDF files provided by the GRC are for personal use only and may not be reproduced. The files reflect the holdings of the GRC library and only contain pages relevant to giraffe study, and may not be complete. Users are obliged to follow all copyright restrictions.