Scholarlink

The vital role of agriculture in global economies and the sustainable development of societies is indisputable. In India, this importance is amplified as the agricultural sector serves as the economic backbone and lifeline for rural populations. This research delves into the realm of crop identification within Maharashtra’s agricultural landscape, with a specific focus on Sugarcane crops. By harnessing the power of Sentinel 2 satellite imagery and employing two popular machine learning algorithms—Support Vector Machines (SVM) and Random Forest (RF), the study aims to achieve precise and efficient crop classification. Through the integration of spectral signatures and Normalized Difference Vegetation Index (NDVI) values, the proposed machine learning models demonstrate an accuracy of around 83% in distinguishing Sugarcane crops. Extensive optimization and tuning are carried out to attain maximum classification precision, resulting in the successful classification of 50,000 randomly generated Ground Control Points (GCPs). By leveraging advanced technologies, this research underscores the potential of remote sensing and machine learning in revolutionizing crop monitoring, enhancing decision-making, and ensuring sustainable agricultural practices, especially within the context of sugarcane cultivation.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning-Based Approach for Sugarcane Crop Classification Using Spectral Signature from Sentinel-2 Data in Maharashtra, India

  • 章节

  • 发表日期：

    2026年

查看全文
  • Varad Purandare,
  • Vidya Kumbhar,
  • Sahil K. Shah,
  • T. P. Singh,
  • Shilpa Gite,
  • Biswajeet Pradhan

摘要

The vital role of agriculture in global economies and the sustainable development of societies is indisputable. In India, this importance is amplified as the agricultural sector serves as the economic backbone and lifeline for rural populations. This research delves into the realm of crop identification within Maharashtra’s agricultural landscape, with a specific focus on Sugarcane crops. By harnessing the power of Sentinel 2 satellite imagery and employing two popular machine learning algorithms—Support Vector Machines (SVM) and Random Forest (RF), the study aims to achieve precise and efficient crop classification. Through the integration of spectral signatures and Normalized Difference Vegetation Index (NDVI) values, the proposed machine learning models demonstrate an accuracy of around 83% in distinguishing Sugarcane crops. Extensive optimization and tuning are carried out to attain maximum classification precision, resulting in the successful classification of 50,000 randomly generated Ground Control Points (GCPs). By leveraging advanced technologies, this research underscores the potential of remote sensing and machine learning in revolutionizing crop monitoring, enhancing decision-making, and ensuring sustainable agricultural practices, especially within the context of sugarcane cultivation.