@inproceedings{62041cca865f471d8bace739be22cc47,
title = "Terahertz Imaging and Machine Learning in the Classification of Coffee Beans",
abstract = "The geographical origin of coffee beans represents an effect on the attributes and quality of the product due to the different soil and weather conditions for a specific location. Therefore, the development of methods for rapid classification and authentication of coffee beans based on their geographical origin is essential. This research was done with the purpose of determining the capacity of coffee (Coffea arabica) varieties classification with the use of Terahertz (THz) imaging and machine learning. THz images of coffee beans samples from 3 different geographical origins were acquired with a time-domain spectrometer and then used to measure the classification performance of methods such as neural networks, random forests, and support vector machines. The results obtained reached an accuracy up to 91.2\%, which showed that the use of THz imaging and machine learning is an effective method for the non-destructive analysis of coffee variables and classification based on geographical origin.",
keywords = "Chemometrics, Coffee post-harvest, Supervised learning, THz spectroscopy",
author = "Patricia Uceda and Hideaki Yoshida and Pedro Castillo",
note = "Publisher Copyright: {\textcopyright} 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 6th Brazilian Technology Symposium, BTSym 2020 ; Conference date: 26-10-2020 Through 28-10-2020",
year = "2021",
doi = "10.1007/978-3-030-75680-2\_94",
language = "English",
isbn = "9783030756796",
series = "Smart Innovation, Systems and Technologies",
pages = "854--861",
editor = "Yuzo Iano and Osamu Saotome and Guillermo Kemper and \{Mendes de Seixas\}, \{Ana Claudia\} and \{Gomes de Oliveira\}, Gabriel",
booktitle = "Proceedings of the 6th Brazilian Technology Symposium, BTSym 2020 - Emerging Trends and Challenges in Technology",
}