Mapping Cropland Extent in Pakistan Using Machine Learning Algorithms on Google Earth Engine Cloud Computing Framework

An actual cropland extent product with a high spatial resolution precision of up to 60 m is believed be particularly significant in tackling numerous water security concerns and world food challenges. To advance the development niche, advanced goods such as crop variety techniques, intensities, production, irrigation, it necessary examine how products typically span narrow or expansive farmlands. Some existing challenges are processing by constructing precision-high cropland-wide items training testing data on diverse geographical locations safe frontiers, computing capacity, managing vast volumes data. This analysis includes eight separate Sentinel-2 multi-spectral instruments from 2018 2019 (Short-wave Infrared Imagery (SWIR 2), SWIR 1, Cirrus, near infrared, red, green, blue, aerosols) have been used. Pixel-based classification algorithms employed, their measured scrutinized this study. The computations analyses conducted cloud-based Google Earth Engine network. Training were obtained map console at 10 for analysis. basis research information computer consists 855 samples, culminating manufacturing field 200 individual validation samples measuring accuracy. Pakistan produced study using four state-of-the-art machine learning (ML) approaches, Random Forest, SVM, Naïve Bayes & CART shows an overall accuracy 82%, 89% manufacturer accuracy, 77% customer Among these algorithms, algorithm overperformed other three, impressive 93%. Pakistan’s average areas calculated 370,200 m2, cropland’s scale indicated that sub-national croplands could measured. offers conceptual change maps utilizing remote sensing multi-date.