Eastern Canada Flocks: Images and manually annotated bird positions ...

• Main Authors: Cruz, Marcos, González-Villa, Javier, Lefebvre, Josée, Gilliland, Scott, St-Pierre, Francis, English, Matthew, Lepage, Christine
• Format: Dataset
• Language: English
• Published: Dryad 2021
• Subjects: FOS Biological sciences; automated image analysis; Wildlife Management; computer vision algorithms; stereology; population census; population size estimation; Machine learning; Deep convolutional neural network; Canada; Common Eider; Somateria mollissima
• Online Access: https://dx.doi.org/10.5061/dryad.98sf7m0hx; https://datadryad.org/stash/dataset/doi:10.5061/dryad.98sf7m0hx

The Eastern Canada (ECA) Flocks data set consists of manually annotated Images from the Common Eider (COEI, Somateria mollissima) Winter Survey and the Greater Snow Geese (GSGO, Anser caerulescens atlanticus) Spring Survey. The images were taken in Eastern Canada using fixed-wing aircraft and manually annotated with ImageJ’s Cell counter plugins. We selected and annotated the ECA Flocks images in order to test the precision of the CountEm flock size estimation method. ECA Flocks includes 179 COEI and 99 GSGO single flock images. We cut each image manually to a rectangle that excluded large parts of the image with no birds. Both versions (original and cut) of each image are available in the data set. We manually annotated 637,555 (124,309 COEI and 514,235 GSGO) bird positions in the cut images from both surveys. Each bird has an associated “Type” which refers to species and/or sex. Sex identification was only possible for adult common eiders since females and immature males are brown birds whereas adult males ... : See ECA_Flocks_README.txt ...