Wildlife big data: Processing images from large scale photo surveys with eMammal

Modern motion sensitive cameras, or camera traps, are non-invasive survey tools that produce a photo “voucher” for each observation and can be used at large scales relatively cheaply and easily. However, they also produce a flood of data and images that can overwhelm researchers. To harness the power of big data for wildlife the Smithsonian has built a workflow that includes custom software, computer vision, cloud based workflows, and expert data review. The process quickly moves photos and data from the field to the eMammal repository and aggregates data from multiple projects to be used for research and conservation.