Google’s Cloud Computing Platform-Based Performance Assessment of Machine Learning Algorithms for Precisely Maize Crop Mapping Using Integrated Satellite Data of Sentinel-2A/B and Planetscope

Image classification is an essential factor for crop mapping and identification. In recent years, many researchers focused on improving data mining/machine learning algorithms to more accurately deal with image classification and predictive problems. The publicly availability of geospatial datasets and free access to cloud-based geo-computing platforms such as Google Earth Engine (GEE) are widely being used for robust mapping and monitoring of crop phenology, acreage estimation and crop yield forecasting. In the present study, the maize (Zea mays) crop has been identified and acreage estimated using integrated Sentinel-2A/B and PlanetScope satellite data for the Rabi/ winter season of 2022 in the Indo-Gangetic Plain. In which, we have assessed and compared the performance of classification and regression trees (CART), support vector machine (SVM) and random forest (RF) algorithms of machine learning (ML) for acreage estimation of maize crops using Google’s GEE cloud computing platform. Wherein, we found that RF outperforms CART and SVM algorithms in the GEE platform with PlanetScope data (90.17% Overall Accuracy (OA) with Kappa 0.89) and also with the integration of PlanetScope and Sentinel-2A/B data (OA = 95.53%, Kappa 0.91). But, CART outperforms RF and SVM algorithms with Sentiel-2A/B data (OA = 88.59%, Kappa 0.85). We have also developed a web-based JavaScript code that can be tested anywhere in the world for robust mapping of crops under various climatic conditions. We expected that this study will be helpful for crop cultivation management, precision agriculture, crop insurance and for making a decision support system to prioritise the input subsidy for farmers.