Expert-defined fuzzy model parameters for UAV-based coffee canopy screening

This record documents the expert-defined fuzzy inference model parameters developed for UAV-based canopy screening in coffee cultivation. The model was developed within the postdoctoral research of Gustavo Di Chiacchio Faulin at São Paulo State University (UNESP), School of Sciences and Engineering, Tupã, Brazil, under the supervision of Luís Roberto Almeida Gabriel Filho (SISPROPe-PD project No. 5896).

The accompanying Excel workbook contains five worksheets: MODEL_SPECIFICATION, MEMBERSHIP_FUNCTIONS, FUZZY_RULES_25, SINGLETON_OUTPUTS, and IPML_CLASS_BANDS. It specifies a zero-order Takagi–Sugeno fuzzy inference model with two normalized inputs, canopy occupancy (IOP) and the Integrated Multispectral Vigor Index (IIVM), both represented on a 0–100 scale.

Each input has five linguistic membership functions. The 25 expert-defined rules combine the inputs using the product of membership degrees, and the Localized Management Priority Index (IPML) is obtained through a weighted average of singleton consequents with values of 10, 30, 50, 70, and 90.

This deposit is limited to the fuzzy-model specification. It does not include original UAV imagery, field-level measurements, spatial geometries, individual analytical observations, or executable source code. The spreadsheet does not reproduce the full remote-sensing, multivariate, clustering, anomaly-detection, or multimodel analysis pipeline reported in the associated research manuscript.

IPML represents a relative expert-defined priority for targeted inspection. It is not an externally validated nutrient diagnosis, fertilizer prescription, or agronomic management recommendation.

The associated manuscript is entitled "Interpretable multimodel triangulation of UAV-derived canopy occupancy and multispectral structure for site-specific screening in coffee" and is intended for submission to Precision Agriculture.