Machine Learning Techniques for Phenology Assessment of Sugarcane Using Conjunctive SAR and Optical Data

Crop phenology monitoring is a necessary action for precision agriculture. Sentinel-1 and Sentinel-2 satellites provide us with the opportunity to monitor crop at high spatial resolution accuracy. The main objective of this study was examine potential data their combination sugarcane phenological stages evaluate temporal behaviour parameters indices. Seven machine learning models, namely logistic regression, decision tree, random forest, artificial neural network, support vector machine, naïve Bayes, fuzzy rule based systems, were implemented, predictive performance compared. Accuracy, precision, specificity, sensitivity or recall, F score, area under curve receiver operating characteristic kappa value used as metrics. research carried out in Indo-Gangetic alluvial plains districts Hisar Jind, Haryana, India. backscatters VV, alpha anisotropy and, among indices, normalized difference vegetation index weighted found be most important features predicting phenology. accuracy models ranged from 40 60%, 56 84% 76 88% data, combined respectively. Area ROC values also supported supremacy use data. This infers that are more efficient than alone.