Desenvolvimento de um software de rastreamento de roedores em arenas

Abstract

Data collection is an important part of academic research, no matter the field, and, therefore, the automation of this task is a reoccurring theme in academia. This thesis aims to automate the acquisition of movement data of rodents during behavioral experiments. Often, research groups that perform this type of experiments, do that manually, subjecting themselves to errors. As to do this acquisition automatically, this thesis proposes the development of a software that analyses the recordings of behavioral experiments. And, utilizing a variety of digital image processing and computer vision techniques, acquires the data. After the development, several validation tests were done. The software presented good results, even in poor lighting conditions. This software is of grate relevance to behavioral experiments in rodents, because it allows precise data acquisition in these experiments.