Deep Learning and Synthetic Imagery for Migratory Bird Species Identification Using Drones

In this thesis work, the use of UAS in the study of migratory shorebird species in Canada is explored with the development of computer vision applications. A deep learning classification model is trained to identify the presence of birds of a given species in an image. Images were collected from UAS for the development of the vision models, and realistic models of the species of interest were used. To address a data scarcity issue, the datasets used were augmented with synthetic data with realistic models of the birds. For evaluation of the quality of the artificially generated images, a novel measure is developed. The synthetic image quality measure showed better results in controlled environments when compared to a popular alternative in the literature. The classifiers trained with the augmented dataset showed appropriate performance, with mean accuracy and standard deviation of 94% +- 0.04 in the test set.