UAV-Based Digital Phenotyping for Quantification of Cotton Jassid (Amrasca biguttula) Injury and Identification of Putatively Resistant Cotton Lines

This repository contains processed data, images, and analysis scripts supporting the associated study of unmanned aerial vehicle (UAV)-based phenotyping for quantifying cotton jassid (Amrasca biguttula) injury severity. The repository is organized into two primary data packages: 1. UAV-derived phenotypic data and statistical analysisA ZIP archive containing the processed UAV-derived datasets used for statistical analyses, along with the R script used to perform the analyses. The datasets include vegetation metrics generated from combinations of vegetation indices, image segmentation methods, and summary statistics. These data were used for correlation analyses, predictor evaluation, multivariate regression, mixed-effects modeling, and identification of cotton lines exhibiting putative resistance to jassid-associated injury. The directory structure is preserved so the analysis can be run using the included relative file paths after extraction. 2. Quantitative segmentation accuracy analysisA ZIP archive containing the images and Python scripts used to quantitatively evaluate the accuracy of the image segmentation approaches examined in the study. The included workflow calculates standard segmentation performance metrics, including Precision, Recall, Accuracy, Specificity, F1 score, and mean Intersection over Union (mIoU), and generates the quantitative results used to compare segmentation performance. The directory structure and required files are preserved so the workflow can be run after extraction. Across the study, 241 UAV-derived predictors were evaluated from combinations of vegetation indices, segmentation methods, and summary statistics. These materials support both the evaluation of UAV-derived predictors against manually assigned injury ratings and the quantitative comparison of image segmentation approaches. The repository contains processed, analysis-ready data and reproducible computational workflows rather than the complete collection of raw UAV imagery. These materials are provided to support transparency and reproducibility of the analyses reported in the associated publication.