TY - GEN
T1 - Accuracy Comparison Between Deep Learning Models for Mexican Lemon Classification
AU - Hernández, Angel
AU - Ornelas-Rodríguez, Francisco Javier
AU - Hurtado-Ramos, Juan B.
AU - González-Barbosa, José Joel
N1 - Publisher Copyright:
© 2021, Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - This paper presents a performance comparison between 8 deep learning models trained to classify Mexican lemons by their visual appearance. The models were trained using 913 lemon images. These images were divided into two classes: faulty and healthy. Half of the models were designed to take color images as input. The other half will take grayscale images. Also, two distributions were used for the training stage. The models were tested against new data, and their performance was acceptable. The best model achieved an accuracy of 92% for the training stage and, for the new data, it was able to classify all the new images correctly.
AB - This paper presents a performance comparison between 8 deep learning models trained to classify Mexican lemons by their visual appearance. The models were trained using 913 lemon images. These images were divided into two classes: faulty and healthy. Half of the models were designed to take color images as input. The other half will take grayscale images. Also, two distributions were used for the training stage. The models were tested against new data, and their performance was acceptable. The best model achieved an accuracy of 92% for the training stage and, for the new data, it was able to classify all the new images correctly.
KW - Citrus quality classification
KW - Computer vision system
KW - Deep learning
KW - Mexican lemon quality
UR - http://www.scopus.com/inward/record.url?scp=85119893226&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-89586-0_5
DO - 10.1007/978-3-030-89586-0_5
M3 - Contribución a la conferencia
AN - SCOPUS:85119893226
SN - 9783030895853
T3 - Communications in Computer and Information Science
SP - 62
EP - 73
BT - Telematics and Computing - 10th International Congress, WITCOM 2021, Proceedings
A2 - Mata-Rivera, Miguel Félix
A2 - Zagal-Flores, Roberto
PB - Springer Science and Business Media Deutschland GmbH
T2 - 10th International Congress on Telematics and Computing, WITCOM 2021
Y2 - 8 November 2021 through 12 November 2021
ER -