Author(s): FURUYA, D. E. G.; BOLFE, E. L.; PARREIRAS, T. C.; BARBEDO, J. G. A.; SANTOS, T. T.; GEBLER, L.

Summary: Fruit growing is important in the global agricultural economy, contributing significantly to food security, job creation, and rural development. With the advancement of technologies, mapping fruits using remote sensing and machine learning (ML) and deep learning (DL) techniques has become an essential tool to optimize production, monitor crop health, and predict harvests with greater accuracy. This study was developed in four main stages. In the first stage, a comprehensive review of the existing literature was made from July 2018 (first article found) to June 2024, totaling 117 articles. In the second stage, a general analysis of the data obtained was made, such as the identification of the most studied fruits with the techniques of interest. In the third stage, a more in-depth analysis was made focusing on apples and grapes, with 27 and 30 articles, respectively. The analysis included the use of remote sensing (orbital and proximal) imagery and ML/DL algorithms to map crop areas, detect diseases, and monitor crop development, among other analyses. The fourth stage shows the data’s potential application in a Southern Brazilian region, known for apple and grape production. This study demonstrates how the integration of modern technologies can transform fruit farming, promoting more sustainable and efficient agriculture through remote sensing and artificial intelligence technologies.

Publication year: 2024

Types of publication: Journal article

Unit: Embrapa Digital Agriculture

Keywords: Agricultura digital, Apple, Aprendizado de máquina, Aprendizado profundo, Deep learning, Digital agriculture, Grape, Inteligência artificial, Machine learning, Maçã, Orchard, Pomar, Sensoriamento Remoto, Uva