Deep Learning aplicado ao reconhecimento de imagem para preservação de animais silvestres: uma pesquisa exploratória.

Abstract

The preservation of wildlife is indispensable to maintain the balance of the ecosystem. Deep Learning (DL) is a recent resource that has been used to support this task. Given the above, this research has as a general objective to present the resources needed to perform image recognition through DL, in order to identify, catalog, and account for wild animals. To this end, an exploratory bibliographic research with a qualitative approach was carried out. The present work presents, therefore, studies that indicate how DL is used in the identification, cataloging, and accounting of animals. From this, it is concluded that applying image recognition through DL to automate the mentioned tasks is advantageous mainly due to the reduction of time and human effort saved.