Sustainable Agriculture Systems and Technologies

16Artificial Machine Learning–Based Classification of Land Cover and Crop Types Using Sentinel‐2A Imagery

Ram Kumar Singh¹, Pavan Kumar², Manoj Kumar³, Keshav Tyagi⁴, and Harshi Jain⁴

¹ Department of Natural Resources, TERI School of Advanced Studies, New Delhi, India

² Department of Forest Biology and Tree Improvement, College of Horticulture and Forestry, Rani Lakshmi Bai Central Agricultural University, Jhansi, Uttar Pradesh, India

³ GIS Centre, Forest Research Institute (FRI), Dehradun, Uttarakhand, India

⁴ Forest Research Institute Deemed to be University (FRIDU), Dehradun, Uttarakhand, India

CONTENTS

• 16.1 Introduction
  • 16.1.1 Random Forest Classifier
• 16.2 Methodology
  • 16.2.1 Data Preprocessing
  • 16.2.2 Random Forest Model
• 16.3 Accuracy Assessment
• 16.4 Results and Discussion
  • 16.4.1 Data Processing
  • 16.4.2 Random Forest Machine Learning
  • 16.4.3 Assessment of Land Cover Accuracy and Its Consistency
• 16.5 Conclusion
• Acknowledgments
• References

16.1 Introduction

The remote sensing data availability pertains to the use of images for various application including identification, mapping, classification, and time‐based changes (Joshi et al. 2006; Singh et al. 2020b). This decade will also be known for the availability of big voluminous time‐series data of remote sensing. This opens a wide range of unprecedented challenges to accurately mapping (Singh et al. 2020a), automate processing (Sun et al. 2013) and modeling (Faivre et al. 2009; Kumar et al. 2019b; Singh et al. ...