Assessment of Hemileia vastatrix Severity in Coffea arabica L., using Image Analysis (Pliman)

Abstract

Objective: To determine the severity index of Hemileia vastatrix on coffee leaves using the Pliman package in R Studio, aiming to optimize quantitative evaluations and improving disease monitoring.

Methodology: Sampling was conducted at Los Barreales farm located in Teocelo, Veracruz, where leaves from the Catuaí amarillo coffee variety, both infected by Hemileia vastatrix and healthy, were selected. The leaves were photographed to capture the leaf surface, and the images were subsequently processed in Photoshop to calculate the healthy and rust-affected areas. Subsequently, the RGB color index of the images was analyzed using the Pliman package in R Studio. Various indices were evaluated, and the NGRDI was selected for the automatic quantification of affected and healthy areas. Finally, an analysis of variance (ANOVA), followed by Tukey's test, was performed to compare significant differences among the samples.

Results: Indices based on specific combinations of RGB colors effectively highlighted subtle differences in leaf reflectance, facilitating the detection of disease symptoms, particularly with the NGRDI index.

Study Limitations: While reflectance spectroscopy is highly accurate, it can be expensive and requires specialized equipment.

Conclusions: Computational tools offer precise and rapid disease detection, providing critical support for integrated pest and disease management strategies in agriculture.