Uso de deep learning para detecção e classificação de frutos de acerola em imagens digitais

Abstract

The agroindustry is constantly faced with the need to intensify and improve its production process. In recent years, in Ceará, the scenario is no different. The generation of large acerola farms in the state has intensified the search for new technologies that can enhance their production. More specifically, when it comes to acerola, the large amount of vitamin C present in the fruits is a significant factor for the growth of its production. With this increase in demand, it is necessary to expand manufacturing and thus seek better conditions for fruit production. With the aim of collaborating with the state of the art on the application of computational technologies in acerola crops, this work proposes the application of Deep Learning to create models for detection of acerola fruits in digital images of acerola trees, as well as the classification of these fruits according to their color.