Coffee Leaf Rust Dataset

Citation Author(s):: David Velasquez (Universidad EAFIT)

Alejandro Sánchez (Universidad EAFIT)

Sebastián Sarmiento (Universidad EAFIT)

Camilo Velásquez (Universidad EAFIT)

Mauricio Toro (Universidad EAFIT)

Edwin Montoya (Universidad EAFIT)

Mikel Maiza (Vicomtech)

Basilio Sierra (University of the Basque Country)

Helmuth Trefftz (Universidad EAFIT)
Submitted by:: David Velasquez
Last updated:: Fri, 07/17/2020 - 10:09
DOI:: 10.21227/pz2n-y676
Data Format:: *.json; *.jpg; zip
Links:: Github

3938 views

Categories:

Keywords:

Coffee Leaf Rust; Multiespectral Data; Plant Disease

ACCESS DATASET CITE

Abstract

Three month Coffee Leaf Rust dataset generated by the Cyber Physical Data Collection System.

Instructions:

The data folder contains five folders (0, 1, 2, 3, 4) previously labeled by the Universidad EAFIT Biology Team, indicating the severity of the Coffee Leaf Rust development stage. The CLR development stage corresponds to the following figure:

De Melo Virginio Filho, E.; Astorga, C. Prevención y Control de la Roya del Café: Manual de Buenas Prácticas para Técnicos y Facilitadores, 1st ed.; CATIE: Turrialba, Costa Rica, 2015; p. 67.

Inside each tagged folder, there are multiple folders with the name lot_# corresponding to a # sample of a lot. This lot folder contains inside the following folders and files:

rgb_images: folder containing the RGB images of the corresponding plants. This folder includes another five sub-folders (0, 1, 2, 3, 4) with the CLR development stage label for each plant used to train a separate RGB model.
lot_#.json: file containing all the sensor information from the # lot.
re.jpg: Multispectral Camera with Red Edge filter picture of the corresponding lot.
rgn.jpg: Multispectral Camera with Red+Green+NIR picture of the corresponding lot.