Identification of Coffee Borer and Rust in Coffee Plants Using Digital Image Processing and Convolutional Neural Networks

• Autor(a) principal: BRAGA VENTURA , CHARLY
• Data de Publicação: 2024
• Outros Autores: OLIVEIRA RODRIGUES, ÉRICK, ANDALÓ MENDES DE CARVALHO, VANESSA, CARVALHO IZIDORO, SANDRO
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Revista de Estudos Interdisciplinares
• Texto Completo: https://revistas.ceeinter.com.br/revistadeestudosinterdisciplinar/article/view/972
• topic: Deep Learning; Visão Computacional; Rede Neural Convolucional; Leucoptera Coffeella; Hemileia Vastatrix; Inteligência artificial; Deep Learning; Computer Vision; Convolutional Neural Network; Leucoptera Coffeella; Hemileia Vastatrix; Artificial Intelligence; Aprendizaje profundo; Visión por computador; Red Neuronal Convolucional; Leucoptera Coffeella; Hemileia Vastatrix; Inteligencia Artificial

The global demand for coffee increases every year, and Brazil is one of the largest producers worldwide. With the high volume of production, there is a growing need to improve the quality of the product due to the demands of both national and international markets. However, pests such as the coffee borer (Leucoptera coffeella) and rust (Hemileia vastatrix) cause significant damage to coffee plantations, leading to annual crop losses. Several methods and techniques have been developed and applied to assess infestation levels and control these pests. Among these techniques are the use of computer vision and convolutional neural networks (CNN). Thus, the goal of this work was to develop a computational tool to correctly identify the presence of pests, reducing evaluation time, the evaluator's error, and labor costs. The accuracy of the developed methods ranged from 99.67% to 97.00%.