Image dataset - Evaluating CNN Models and Optimization Techniques for Quality Classification of Dried Chili Peppers (Capsicum annuum L.)

Published: 28 April 2026| Version 1 | DOI: 10.17632/vrby4t2w57.1
Contributor:
Carlos Guerrero-Mendez

Description

Dataset of Dried Chili Pepper (Capsicum annuum L.) Images for Quality Classification This dataset contains digital images of dried chili peppers (Capsicum annuum L.), commonly known as "Mirasol" or "Guajillo," collected to support the development and evaluation of automated visual classification systems for sorting machines in the agricultural industry. Images are organized into five categories aligned with the Mexican Official Norm (NMX-FF-107/1-SCFI-2014) and operational scenarios expected in an industrial sorting machine: Extra: Highest quality peppers, with uniform color and intact structure. First Class: Good quality peppers with minor imperfections. Second Class: Lower quality peppers, which may present discoloration or minor damage. Trash: Heavily fragmented, damaged, or otherwise non-sellable peppers. Empty: Images with no chili pepper present, representing the empty conveyor scenario common in sorting machine operation. The dataset comprises a total of 2,866 images distributed as follows: 564 Extra, 529 First Class, 611 Second Class, 541 Trash, and 621 Empty. Evaluating CNN Models and Optimization Techniques for Quality Classification of Dried Chili Peppers (Capsicum annuum L.) Lopez-Betancur, Daniela; Saucedo-Anaya, Tonatiuh; Guerrero-Mendez, Carlos; Navarro-Solís, David; Silva-Acosta, Luis; Robles-Guerrero, Antonio; Gomez-Jimenez, Salvador.  International Journal of Combinatorial Optimization Problems and Informatics; Jiutepec Tomo 15, N.º 2, (2024): 13-25. DOI:10.61467/2007.1558.2024.v15i2.462

Files

Institutions

Categories

Computer Vision, Food Quality, Image Classification, Computer in Foodservice Applications, Dried Food, Convolutional Neural Network, Deep Learning, Transfer Learning, Agriculture

Licence