Classification for Land Use and Land Cover

Abstract

In this study, we examine the crucial areas of level 2 categorization and crop discrim- ination in the Indian district of Udham Singh Nagar.Finding the model with the best accuracy in level 2 classification and crop discrimination is the main goal of this study. providing useful information for crop monitoring, estimating productivity, and putting improved agricultural methods into practice. The use of Google Earth Engine, a po- tent platform that facilitates the effective processing and analysis of enormous satellite datasets, is essential to our methodology. We do thorough assessments of the trained models, evaluating their ability to distinguish between various crop types, by utilizing Earth Engine’s capabilities. Crop discrimination and level 2 categorization are the re- search’s focus problems. This indicates that the research uses remote sensing data to detect and classify the particular crops cultivated in various Udham Singh Nagar dis- trict locations. Using this method, the spectral characteristics of the land cover that the satellite photography has captured are immediately analyzed. Crops can be classi- fied using spectral patterns that the models can learn, as different crops reflect sunlight differently at different wavelengths. Data collection, preprocessing, feature extraction, machine learning model training, model evaluation, and result analysis are the steps in this study