Satellite Image Analysis: Clustering and Classification (eBook)

Surekha Borra is currently a professor at the Department of Electronics and Communication Engineering and chief research coordinator of KS Institute of Technology, Bangalore, India. She earned her doctorate in the copyright protection of images from Jawaharlal Nehru Technological University, Hyderabad, India. Her research interests include image and video analytics, machine learning, biometrics, biomedical signals, and remote sensing. She has filed one Indian patent and published seven books, sixteen book chapters, and several research papers in refereed and indexed journals, and in the proceedings of international conferences. She has received several research grants and awards from professional bodies and the Karnataka state government of India, including a Distinguished Educator & Scholar Award for her contributions to teaching and scholarly activities, and a Woman Achiever’s Award from the Institution of Engineers (India) for her prominent and innovative research. Rohit Thanki earned his PhD in multibiometric system security using the compressive sensing theory and watermarking from CU Shah University, Gujarat, India, in 2017. His research interests include digital watermarking, the biometrics system, security, compressive sensing, pattern recognition, and image processing. He has published seven books, seven book chapters, and more than 25 research papers in refereed and indexed journals, and has participated in conferences at the national and international level. He currently serves as a reviewer for journals published by the Institute of Electrical and Electronics Engineers (IEEE), Elsevier, Taylor & Francis, Springer, and IGI Global. Nilanjan Dey is an Assistant Professor at the Department of Information Technology, Techno India College of Technology, Kolkata, India. He was an honorary Visiting Scientist at Global Biomedical Technologies Inc., CA, USA and an associated Member of the University of Reading, London, UK. He has authored or edited more than 40 books with Elsevier, Wiley, CRC Press and Springer etc., and published more than 300 research articles.  He is Editor-in-Chief of the International Journal of Ambient Computing and Intelligence, IGI Global, USA. He is the Series Co-Editor of Springer Tracts in Nature-inspired Computing; Advances in Ubiquitous Sensing Applications for Healthcare; and Intelligent Signal Processing and Data Analysis; as well as an Associated Editor for IEEE Access. His main research interests include medical imaging, machine learning, data mining etc.  Recently, he was selected as one of the top 10 most published and cited academics in the field of Computer Science in India during the period of consideration 2015-17. 

About the Book 6
Contents 7
About the Authors 10
List of Figures 12
List of Tables 14
1 Introduction 16
1.1 Introduction 16
1.2 Satellite Imaging Sensors 17
1.3 Panchromatic and Multispectral Images 17
1.4 Resolution in Satellite Images 19
1.5 Distortions in Satellite Images 19
1.6 Manual Versus Automatic Interpretation 20
1.7 Classification and Clustering 21
1.8 Performance Evaluation of Classification Techniques 22
1.9 Conclusion 26
References 26
2 Satellite Image Enhancement and Analysis 28
2.1 Satellite Image Degradation and Restoration 28
2.2 Geometric Correction or Rectification in Satellite Images 28
2.3 Noise Removal 30
2.4 Satellite Image Enhancement 30
2.5 Satellite Image Segmentation 34
2.6 Image Stitching 37
2.7 Satellite Image Interpolation 38
2.8 Multivariate Image Processing 39
2.9 Image Differencing 40
2.10 Band Ratioing 40
2.11 Other Image Transformations 41
References 43
3 Satellite Image Clustering 45
3.1 Introduction 45
3.2 Supervised Classification 47
3.3 Unsupervised Classification (Clustering) 48
3.4 K-means Clustering 50
3.5 Iterative Self-organizing Data Analysis (ISODATA) 50
3.6 Gaussian Mixture Models 52
3.7 Self-organizing Maps 54
3.8 Hidden Markov Models 56
3.9 Feature Extraction and Dimensionality Reduction 58
3.10 Conclusion 61
References 62
4 Satellite Image Classification 67
4.1 Introduction 67
4.2 Supervised Classification 67
4.3 Max Likelihood Classifier 70
4.4 Naïve Bayes 72
4.5 K-Nearest Neighbors (KNN) 74
4.6 Minimum Distance to Means (MDM) 76
4.7 Parallelepiped Classifier 77
4.8 Support Vector Machine (SVM) 78
4.9 Discriminant Analysis (DA) 81
4.10 Decision Trees 82
4.11 Binary Encoding Classification 84
4.12 Spectral Angle Mapper Classification 85
4.13 Artificial Neural Network (ANN) 86
4.14 Deep Learning (DL) 86
4.15 The Hybrid Approaches 88
4.16 Semi-supervised Learning 89
4.17 Challenges 90
References 91
5 Applied Examples 96
5.1 Introduction 96
5.2 Agriculture 97
5.3 Forestry 98
5.4 Rainfall Estimation 98
5.5 Disaster Monitoring and Emergency Mapping 99
5.6 Biodiversity 100
5.7 Epidemiological Study 101
5.8 Oceanography 102
5.9 Maritime/Illegal Fishing 102
5.10 Coastal Zone Management 102
5.11 Road Detection 103
5.12 Vehicle Detection 104
5.13 Aircraft Detection 105
5.14 Thermal Applications 105
5.15 Meteorology 105
5.16 Heritage Management 106
5.17 Challenges and Future Perspectives 106
References 106