Machine Learning Algorithms for Signal and Image Processing E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 1097

Veröffentlichungsjahr: 2022

Beschreibung

Machine Learning Algorithms for Signal and Image Processing Enables readers to understand the fundamental concepts of machine and deep learning techniques with interactive, real-life applications within signal and image processing Machine Learning Algorithms for Signal and Image Processing aids the reader in designing and developing real-world applications using advances in machine learning to aid and enhance speech signal processing, image processing, computer vision, biomedical signal processing, adaptive filtering, and text processing. It includes signal processing techniques applied for pre-processing, feature extraction, source separation, or data decompositions to achieve machine learning tasks. Written by well-qualified authors and contributed to by a team of experts within the field, the work covers a wide range of important topics, such as: * Speech recognition, image reconstruction, object classification and detection, and text processing * Healthcare monitoring, biomedical systems, and green energy * How various machine and deep learning techniques can improve accuracy, precision rate recall rate, and processing time * Real applications and examples, including smart sign language recognition, fake news detection in social media, structural damage prediction, and epileptic seizure detection Professionals within the field of signal and image processing seeking to adapt their work further will find immense value in this easy-to-understand yet extremely comprehensive reference work. It is also a worthy resource for students and researchers in related fields who are looking to thoroughly understand the historical and recent developments that have been made in the field.

Leseprobe

Table of Contents

Cover

Title Page

Copyright

Editor Biography

List of Contributors

Preface

Acknowledgments

Section I: Machine Learning and Deep Learning Techniques for Image Processing

1 Image Features in Machine Learning

1.1 Introduction

1.2 Feature Vector

1.3 Lower‐Level Image Features

1.4 Conclusion

References

2 Image Segmentation and Classification Using Deep Learning

2.1 Introduction

2.2 Image Segmentation

2.3 Image Classification

2.4 Conclusion

References

3 Deep Learning Based Synthetic Aperture Radar Image Classification

3.1 Introduction

3.2 Literature Review

3.3 Dataset Description

3.4 Methodology

3.5 Experimental Results and Discussions

3.6 Conclusion

References

4 Design Perspectives of Multi‐task Deep‐Learning Models and Applications

4.1 Introduction

4.2 Deep Learning

4.3 Multi‐task Deep‐Learning Models

4.4 Design and Implementation

4.5 Applications

4.6 Evaluation of Multi‐task Models

4.7 Conclusion and Future Directions

Acknowledgment

References

5 Image Reconstruction Using Deep Learning

5.1 Introduction

5.2 DL‐IR Methods

5.3 DL‐Based Medical Image Reconstruction

5.4 Conclusion

Acknowledgment

References

6 Machine and Deep‐Learning Techniques for Image Super‐Resolution

6.1 Introduction

6.2 Traditional Upsampling Approaches

6.3 Primitive Machine‐Learning‐Based Approaches

6.4 Modern Deep‐Learning‐Based Approaches

6.5 Performance Metrics and Comparative Study of Existing Techniques

6.6 Summary and Discussions

References

Section II: Machine Learning and Deep Learning Techniques for Text and Speech Processing

7 Machine and Deep‐Learning Techniques for Text and Speech Processing

7.1 Text Processing

7.2 Speech Processing

7.3 Conclusion

References

8 Manipuri Handwritten Script Recognition Using Machine and Deep Learning

8.1 Introduction

8.2 Literature Survey

8.3 Proposed Work

8.4 Experimental Results and Discussions

8.5 Conclusion

References

9 Comparison of Different Text Extraction Techniques for Complex Color Images

9.1 Introduction

9.2 Related Work

9.3 Edge‐Based and CC‐Based Methods

9.4 Proposed Methodology

9.5 Experimental Results and Discussion

9.6 Conclusions

Acknowledgment

References

10 Smart Text Reader System for People who are Blind Using Machine and Deep Learning

10.1 Introduction

10.2 Literature Review

10.3 Experimental Results

10.4 Conclusions and Recommended Future Work

Acknowledgments

References

Notes

11 Machine‐Learning Techniques for Deaf People

11.1 Introduction

11.2 Literature Survey

11.3 Objectives

11.4 Proposed Calculation Depiction

11.5 Resources and Strategies

11.6 Assessment

11.7 Outcomes and Conversations

11.8 Discourse Coherence

11.9 Conclusion

References

12 Design and Development of Chatbot Based on Reinforcement Learning

12.1 Introduction

12.2 Student Guide Using Chatbot

12.3 Implementation of Chatbot System

12.4 Development of Algorithms Used in Chatbot System

12.5 Conclusion

References

13 DNN Based Speech Quality Enhancement and Multi‐speaker Separation for Automatic Speech Recognition System

13.1 Introduction

13.2 Deep Learning

13.3 Speech Enhancement and Separation

13.4 Speech Enhancement Algorithms

13.5 Speech Separation Algorithms

13.6 Deep Learning Based Speech Enhancement

13.7 Deep Learning Based Speech Separation

13.8 Results and Discussions

13.9 Conclusion

References

14 Design and Development of Real‐Time Music Transcription Using Digital Signal Processing

14.1 Introduction

14.2 Related Work

14.3 Motivation of the Proposed Work

14.4 Mathematical Expressions of Signal Processing

14.5 Proposed Methodology

14.6 Experimental Results and Discussions

14.7 Conclusion

References

Section III: Applications of Signal and Image Processing with Machine Learning and Deep Learning Techniques

15 Role of Machine Learning in Wrist Pulse Analysis

15.1 Introduction

15.2 Machine‐Learning Techniques

15.3 Performance Analysis of ML Algorithms

15.4 Role of the Machine and Deep Learning in Wrist Pulse Signal Analysis (WPA)

15.5 Discussion and Conclusion

References

16 An Explainable Convolutional Neural Network‐Based Method for Skin‐Lesion Classification from Dermoscopic Images

16.1 Introduction

16.2 Methods and Materials

16.3 Explainable Deep‐Learning (‐DL) Framework for Dermoscopic Image Classification

16.4 Experimental Results and Discussion

16.5 Conclusion

Acknowledgments

References

17 Future of Machine Learning (ML) and Deep Learning (DL) in Healthcare Monitoring System

17.1 Introduction

17.2 Performance Analysis Parameters

17.3 Objectives and Motivation

17.4 Existing ML/DL Techniques for Healthcare Monitoring and Disease Diagnosis

17.5 Proposed Model/Methods for Healthcare Monitoring System Using ML/DL

17.6 Experimental Results and Discussion

17.7 Conclusions

17.8 Future Scope

References

18 Usage of AI and Wearable IoT Devices for Healthcare Data: A Study

18.1 Introduction

18.2 Literature Review

18.3 AI‐Based Wearable Devices

18.4 Activities of Wearable Devices in Healthcare System

18.5 Barriers to Wearable's Adoption

18.6 Wearable Devices Consumers

18.7 Recent Trends in Wearable Technology

18.8 Conclusion

References

19 Impact of IoT in Biomedical Applications Using Machine and Deep Learning

19.1 Introduction

19.2 History of DL and ML

19.3 Methods of ML and DL Algorithms and Classification

19.4 ML and DL Applications in Biomedicine

19.5 Discussions of IoT‐Based ML and DL Case Studies in Biomedical Systems

19.6 Opportunities and Challenges

References

20 Wireless Communications Using Machine Learning and Deep Learning

20.1 Introduction

20.2 Contributions of Intelligent Reflecting Surfaces (IRS) in Wireless‐Communication Systems

20.3 Merits of IRS‐Aided Wireless‐Communication Systems for Performance Enhancement

20.4 Issues in Collaboration Between Active and Passive Beamforming

20.5 Scope of Machine Learning for IRS‐Enabled Wireless‐Communication Systems

20.6 Summary

Acknowledgment

References

21 Applications of Machine Learning and Deep Learning in Smart Agriculture

21.1 Introduction

21.2 Concept of Machine Learning

21.3 Concept of Deep Learning

21.4 Smart Agriculture

21.5 Computation Methods

21.6 Security Aspects and Issues

21.7 Application Domains in Agriculture

21.8 Case Study

21.9 Agro Smart City

21.10 Concept of Application of ML and DL in Smart Agriculture

21.11 Results and Discussion

21.12 Conclusion

References

22 Structural Damage Prediction from Earthquakes Using Deep Learning

22.1 Introduction

22.2 Literature Review

22.3 Proposed Methodology

22.4 Proposed Methodology for Deep‐Learning Models

22.5 Experimental Results and Discussions

22.6 Conclusion

References

23 Machine‐Learning and Deep‐Learning Techniques in Social Sciences

23.1 Introduction

23.2 Machine‐Learning and Deep‐Learning Techniques

23.3 Social Sciences Applications Using Machine‐Learning and Deep‐Learning Techniques

23.4 Conclusion

References

24 Green Energy Using Machine and Deep Learning

24.1 Introduction

24.2 ML Algorithms for Green Energy

24.3 Managing Renewable‐Energy Integration with Smart Grid

24.4 DL Models for Renewable Energy

24.5 Conclusion

References

25 Light Deep CNN Approach for Multi‐Label Pathology Classification Using Frontal Chest X‐Ray

25.1 Introduction

25.2 Related Work

25.3 Materials and Method

25.4 Proposed Methodology

25.5 Result and Discussions

25.6 Conclusion

References

Index

End User License Agreement

List of Tables

Chapter 3

Table 3.1 Details of MSTAR targets used in this experiment.

Table 3.2 Confusion matrix of SAR target recognition results.

Table 3.3 Evaluation indexes of models with different types of input images...

Chapter 5

Table 5.1 List of AE‐based DL methods for different IR tasks.

Table 5.2 List of CNN‐based DL‐MMSE methods for different IR tasks.

Chapter 6

Table 6.1 Average PSNR/SSIM values for the scale factors 2×, 3×, ...

Chapter 7

Table 7.1 Automatic image caption generation using machine‐ and deep‐learni...

Table 7.2 Recognition of MM handwritten script methods.

Chapter 8

Table 8.1 Confusion matrix of 10 digits of MMM.

Table 8.2 Percentage accuracy for recognition.

Table 8.3 Comparison of performance.

Table 8.4 Comparison between the proposed model with existing models.

Chapter 9

Table 9.1 Size of sliding window.

Table 9.2 Selection of

k

values.

Table 9.3 Datasets description used for experimentation.

Table 9.4 List of evaluation parameters and their formula.

Table 9.5 Number of clusters chosen.

Table 9.6 Comparison between different text extraction techniques (i.e. edg...

Table 9.7 Performance evaluation metrics of different text extraction techn...

Table 9.8 Comparison of proposed method with other techniques for complex c...

Table 9.9 Benefits of the proposed method.

Chapter 10

Table 10.1 Deep learning text detection methods, where W, word; T, text‐lin...

Table 10.2 Supplementary table of abbreviations.

Table 10.3 Comparison among state‐of‐the‐art of deep learning‐based text re...

Table 10.4 Abbreviation descriptions.

Table 10.5 Comparison among some of the recent text detection and recogniti...

Table 10.6 Quantitative comparison among some of the recent text detection ...

Table 10.7 Comparing some of the recent text recognition techniques using W...

Chapter 11

Table 11.1 Proportions of planned convolutional neural network model.

Table 11.2 Segment information.

Table 11.3 Each subject and conditions discourse comprehensibility record e...

Table 11.4 Each subject and conditions discourse comprehensibility record es...

Table 11.5 Each subject and conditions discourse comprehensibility record es...

Table 11.6 All strategies wide‐ranging value and short preparing postpone e...

Chapter 14

Table 14.1 Representation of octaves according to their MIDI pitch ranges....

Table 14.2 Representation pitch classes according to y‐coordinates.

Table 14.3 Representation of obtained key and their features.

Table 14.4 Comparison of the proposed technique with existing techniques.

Chapter 15

Table 15.1 Machine‐learning techniques used for analysis of radial pulse si...

Chapter 16

Table 16.1

x

‐DL (CNN‐2P) architecture details for dermoscopic image classif...

Table 16.2 Architecture and performance metrics (%) of various explainable ...

Chapter 17

Table 17.1 Comparison of technology for disease detection and diagnosis [1–...

Table 17.2 Results comparison of proposed technology with existing technolo...

Chapter 18

Table 18.1 Comparison among the premium wearable devices.

Table 18.2 Global growth rate.

Chapter 19

Table 19.1 Review of the examined ML and DL algorithms.

Table 19.2 Review of the examined ML and DL applications.

Chapter 20

Table 20.1 Contribution of machine learning‐based approaches for IRS‐enable...

Chapter 22

Table 22.1 Experimental results.

Chapter 25

Table 25.1 Related work to chest X‐ray classification.

Table 25.2 MobileNet V2 bottleneck block architecture.

Table 25.3 Our method's experimental results.

Table 25.4 AUC result overview for our experiment.

List of Illustrations

Chapter 1

Figure 1.1 Circular (8, 1) and (16, 2) neighborhoods.

Chapter 2

Figure 2.1 A pixel predictive model based on FCN. g.t, means the ground trut...

Figure 2.2 The DeepLab architecture.

Figure 2.3 The DeepLab‐V3+ model architecture.

Figure 2.4 Deconvolutional segmentation (semantic).

Figure 2.5 The SegNet framework.

Figure 2.6 Mask R‐CNN model architecture.

Figure 2.7 The PSPN model architecture.

Figure 2.8 The ReSeg framework.

Figure 2.9 The graph‐LSTM semantic segmentation model.

Figure 2.10 The GAN‐based semantic segmentation model.

Figure 2.11 Attention‐based model for semantic segmentation.

Figure 2.12 The CNN–RNN pipeline framework.

Figure 2.13 The RNN model for multi‐label classification.

Chapter 3

Figure 3.1 Visualization of SAR image of target: Bulldozer‐D7. (a) 2D magnit...

Figure 3.2 Process flow of three channeled preprocessed image generation.

Figure 3.3 Visual comparison of optical images (normal) and generated SAR (a...

Figure 3.4 Architecture of proposed CNN model with three channel SAR input....

Figure 3.5 Precision–recall curves results of model of different classes.

Chapter 4

Figure 4.1 Single‐task learning versus multi‐task learning diagram. (a) Sing...

Figure 4.2 AI landscape and sub‐emerging fields.

Figure 4.3 Hard parameter (a) and soft parameter (b) sharing MTL using joint...

Figure 4.4 R‐CNN for (a) face detection, (b) landmark locations, (c) pose es...

Figure 4.5 MTL architecture by combination of several R‐CNN.

Figure 4.6 LSTM multi‐task architecture.

Figure 4.7 Network architecture of the split layers.

Figure 4.8 Hard parameter sharing.

Figure 4.9 Soft parameter sharing.

Figure 4.10 Deep relationship network.

Figure 4.11 Fully adaptive feature sharing.

Figure 4.12 Cross‐stitch networks.

Figure 4.13 Weighting losses with uncertainty.

Figure 4.14 Sluice networks.

Figure 4.15 Joint many‐task model.

Chapter 5

Figure 5.1 Schematic representation of image reconstruction in computer visi...

Figure 5.2 (a) Medical image reconstruction (using DL model) from the raw da...

Figure 5.3 Schematic representation of image reconstruction using an autoenc...

Figure 5.4 A generic representation of DL‐IR methods using CNNs. The CNN‐bas...

Figure 5.5 Schematic representation of training and testing a GAN network fo...

Chapter 6

Figure 6.1 Single image super‐resolution framework.

Figure 6.2 Traditional upsampling methods available in GIMP, photo‐editing s...

Figure 6.3 Application of SR on remote sensing data. (a) Bilinearly interpol...

Figure 6.4 Application of SR on brain MRI scan (BrainWeb database). (a) Bicu...

Figure 6.5 Upsampling a 3 × 3 image using nearest‐neighbor interpolation (to...

Figure 6.6 Block diagram of the machine‐learning framework for image super‐r...

Figure 6.7 Upsampling‐based classification of the super‐resolution framework...

Figure 6.8 Linear networks for super‐resolution.

Figure 6.9 Residual networks used to perform super‐resolution.

Figure 6.10 Recursive networks for super‐resolution.

Figure 6.11 Progressive reconstruction networks for super‐resolution.

Figure 6.12 Densely connected networks used in image super‐resolution.

Figure 6.13 Attention‐based networks for super‐resolution.

Figure 6.14 GAN‐based networks for super‐resolution.

Chapter 7

Figure 7.1 The architecture of the automatic text to image system.

Figure 7.2 The methodology of ligature detection, orientation, and recogniti...

Figure 7.3 The block diagram of captioning the image.

Figure 7.4 Block diagram of MM handwritten recognition model.

Figure 7.5 Meitei alphabets and numerals.

Figure 7.6 Block diagram of smart sign language recognition system.

Figure 7.7 Signs for English alphabets can be used as dataset for the system...

Figure 7.8 The methodology stages of smart reader for blind people.

Figure 7.9 The stages of recognition of ASR system.

Chapter 8

Figure 8.1 Meitei Mayek script.

Figure 8.2 Proposed model.

Figure 8.3 An image script sample from MMM numerals dataset.

Figure 8.4 Accuracy of model.

Figure 8.5 Model loss.

Chapter 9

Figure 9.1 Flowchart of edge‐based method.

Figure 9.2 Steps involved in edge‐based method.(a) Input image; (b) Gaus...

Figure 9.3 Flowchart of CC‐based method.

Figure 9.4 Steps involved in CC‐based method formulated.(a) Input image;...

Figure 9.5 Flowchart of proposed text extraction method.

Figure 9.6 Flowchart of

k

‐means clustering algorithm.

Figure 9.7 Text extraction steps for a complicated color image: (a) input im...

Figure 9.8 Images from complex color images before and after text extraction...

Chapter 10

Figure 10.1 Examples for two main classes of text: text in a printed documen...

Figure 10.2 General taxonomy for text detection.

Figure 10.3 General schematic of scene text detection and recognition.

Figure 10.4 Illustrative example for semantic versus instance segmentation. ...

Figure 10.5 Comparison among some of the recent 1D CTC‐based scene text reco...

Figure 10.6 Comparing the processing steps for tackling the character recogn...

Figure 10.7 Comparison of recent attention‐based scene text recognition fram...

Figure 10.8 Sample images of synthetic datasets used for training in scene t...

Figure 10.9 Qualitative detection results comparison among CRAFT...

Figure 10.10 Distribution of word height in pixels computed on the test set ...

Figure 10.11 Evaluation of the text detection performance for CRAFT...

Figure 10.12 Average

H

‐mean versus frames per second (FPS) computed on ICDAR...

Figure 10.13 Qualitative results for challenging examples of scene text reco...

Figure 10.14 Illustration for challenging cases on scene text recognition th...

Figure 10.15 Statistics of word length in characters computed on (a) ICDAR13...

Figure 10.16 Evaluation of the average WRA at different word length for ASTE...

Figure 10.17 Statistics of word aspect‐ratios computed on (a) ICDAR13, (b) I...

Figure 10.18 Evaluation of average WRA at various word aspect‐ratios for AST...

Figure 10.19 Average WRA versus average recognition time per word in millise...

Chapter 11

Figure 11.1 Flowchart of proposed design.

Figure 11.2 Acquired gathering mean level of effectively perceived character...

Figure 11.3 Mean clarity outcome of two examination circumstances: (a) 0 dB ...

Figure 11.4 Case and bristle plans of discourse superiority evaluations for ...

Chapter 12

Figure 12.1 Chatbot system architecture for academics.

Figure 12.2 Data‐flow diagram for handling the queries in academics.

Figure 12.3 Class‐diagram design and development of chatbot based on reinfor...

Figure 12.4 Sequence diagram.

Figure 12.5 Activity diagram of chatbot system.

Figure 12.6 State transition diagram of chat.

Chapter 13

Figure 13.1 RNN architecture.

Figure 13.2 CNN architecture.

Figure 13.3 Speech enhancement algorithms.

Figure 13.4 Speech separation algorithms.

Chapter 14

Figure 14.1 Representation of visualized sample input (in the form of symbol...

Figure 14.2 Flowchart of proposed methodology for music transcription system...

Figure 14.3 Input signal.

Figure 14.4 Scalogram of input signal formed by using CWT.

Figure 14.5 Power spectrogram of input formed by using STFT.

Figure 14.6 Chroma of the input signal using STFT.

Figure 14.7 Steps involved in image‐processing process: (a) cropped image; (...

Figure 14.8 Accuracy for finding pitch class on different types of piano.

Figure 14.9 Beat of music for different types of piano.

Figure 14.10 Tempo of music for different types of piano.

Chapter 15

Figure 15.1 Block diagram of wrist pulse signal processing.

Figure 15.2 Timeline in the development of machine learning and deep learnin...

Figure 15.3 The relationships between AI, ML, and DL.

Figure 15.4 (a) Distribution of various machine‐learning techniques used for...

Chapter 16

Figure 16.1 Four types of dermoscopic images taken from [3–6]: (a) MEL, (b) ...

Figure 16.2 Expert annotations on dermoscopic images: (a), (b) MEL, (c) NEV,...

Figure 16.3 Explainable deep learning (

x

‐DL) framework. The input dermoscopi...

Figure 16.4 Explainable deep learning (

x

‐DL) framework employing CNN with tw...

Figure 16.5 Relevance maps for melanoma skin lesions (MEL) obtained via

x

‐CN...

Figure 16.6 Relevance maps for nevus skin lesions (NEV) obtained via

x

‐CNN‐2...

Figure 16.7 Relevance maps for basal cell carcinoma skin lesions (BCC) obtai...

Figure 16.8 Relevance maps for seborrheic keratoses (SEBK) skin lesions obta...

Figure 16.9 Relevance maps for skin lesions, i.e. MEL (first row), NEV (seco...

Chapter 17

Figure 17.1 Block diagram of proposed system.

Figure 17.2 Categories versus probabilities using traditional NN of random d...

Figure 17.3 Probability of random data on BNN.

Figure 17.4 Probability distribution of random data using 20 binarized neura...

Figure 17.5 Probability (a) and its distribution (b) for four different pred...

Figure 17.6 Probability using NN.

Figure 17.7 Probability (a) and its distribution (b) for four different pred...

Figure 17.8 Random data (a) and Pima diabetes (b) with less confidence.

Figure 17.9 Probability estimate (a) and probability distribution (b).

Figure 17.10 Correlation between samples of diabetes (a) and breast cancer (...

Figure 17.11 Comparison between different techniques based on the accuracy....

Figure 17.12 Confusion matrix for breast cancer (a) and diabetes (b).

Chapter 18

Figure 18.1 Sensor combination inside AVA sensor.

Figure 18.2 Dataflow of AVA sensor.

Figure 18.3 AliveCor – personal EKG.

Figure 18.4 TempTraq.

Figure 18.5 BioScarf.

Figure 18.6 Blinq‐wearable rings.

Figure 18.7 Wearable sleep headband.

Figure 18.8 BioPatches for bone regeneration.

Figure 18.9 Smart glasses.

Figure 18.10 Smart hearing aids: (a) hearing aids, (b) technology used in Li...

Figure 18.11 Wireless patient monitoring: (a) leaf sensor., (b) data flo...

Figure 18.12 Wearable fitness tracker: (a) wristbands, (b) algorithm of wear...

Figure 18.13 (a) Smart health watch., (b) flow diagram of smart health w...

Figure 18.14 ECG monitors.

Figure 18.15 Blood pressure monitor.

Figure 18.16 (a) Biosensor., (b) working principle of biosensor.

Figure 18.17 Percentage of consumers of wearable devices.

Figure 18.18 (a) Impacts on IoT supported wearable devices, (b) impacts on I...

Figure 18.19 Consumer's desirable information.

Figure 18.20 Global wearable shipments in million units.

Figure 18.21 Consumers in India age wise.

Figure 18.22 Consumer details in India year wise ($).

Figure 18.23 Consumer details in global year wise ($).

Figure 18.24 Percentage of global comparison of consumer of wearable device....

Chapter 19

Figure 19.1 Most common neural network structures applied in biomedical: (op...

Figure 19.2 The multilayer perceptron and the clustering layer of the CNN de...

Figure 19.3 Categorization of DL applications for IoT in the medical industr...

Chapter 20

Figure 20.1 (a) cellular communication may affect the human brain; (b) worki...

Figure 20.2 Open‐system interconnection model for communication systems with...

Figure 20.3 Simplified schematic of IRS in a wireless‐communication system....

Chapter 21

Figure 21.1 Basic concept of machine learning.

Figure 21.2 Linear regression.

Figure 21.3 Logistic regression.

Figure 21.4

K

‐nearest neighbor.

Figure 21.5 Example of decision tree.

Figure 21.6 Random forest.

Figure 21.7 Support vector machine.

Figure 21.8 Naïve Bayes method.

Figure 21.9 Clustering.

Figure 21.10 Markov decision process via reinforcement learning.

Figure 21.11 Deep learning.

Figure 21.12 Back propagation.

Figure 21.13 Smart farming.

Figure 21.14 Precision agriculture.

Figure 21.15 ML and DL applications in agriculture.

Figure 21.16 Plant disease classification.

Figure 21.17 Plant classification a) Apple cedar b) Apple scab c) Diseased A...

Figure 21.18 Smart irrigation in agriculture.

Figure 21.19 Pest control treatment.

Figure 21.20 Soil management system.

Chapter 22

Figure 22.1 Flowchart of proposed methodology.

Figure 22.2 1D CNN layers.

Figure 22.3 Spectrogram for 2D CNN.

Figure 22.4 Feature extraction using filters in 2D CNN.

Figure 22.5 2D CNN layers.

Figure 22.6 LSTM layers.

Chapter 23

Figure 23.1 Machine‐learning model.

Figure 23.2 Deep‐learning model.

Figure 23.3 Machine‐learning process.

Chapter 24

Figure 24.1 Advantages of green energy.

Figure 24.2 Classification of solar energy technology.

Figure 24.3 Equivalent circuit of single PV cell.

Figure 24.4 Solar PV cell, PV module, and PV array.

Figure 24.5 Off/on‐grid‐connected PV system.

Figure 24.6 General block diagram of WECS.

Figure 24.7 General block diagram of hydropower plant.

Chapter 25

Figure 25.1 MobileNet V2 bottleneck: The base diagram of the bottleneck bloc...

Figure 25.2 Proposed model architecture: The first block focuses on the data...

Figure 25.3 The Chest X‐ray 14 dataset is represented by six samples. The ut...

Figure 25.4 Class distribution inside the dataset. The distribution of patie...

Figure 25.5 Model loss: The evolution of model Loss curve during training an...

Figure 25.6 Model accuracy: The evolution of model Accuracy during training ...

Figure 25.7 ROC curves for our model across 40 epochs: (AUC 0.91) for “Emphy...

Figure 25.8 Model prediction based on the test set.

Guide

Cover

Table of Contents

Title Page

Copyright

Editor Biography

List of Contributors

Preface

Acknowledgments

Begin Reading

Index

End User License Agreement

Pages

ii

iii

iv

xix

xx

xxi

xxii

xxiii

xxiv

xxv

xxvi

xxvii

xxix

xxx

xxxi

1

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

115

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

195

196

197

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

219

220

221

222

223

224

225

226

227

228

229

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

265

266

267

268

269

270

271

272

273

274

275

276

277

279

280

281

282

283

284

285

286

287

288

289

290

291

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

339

340

341

342

343

344

345

346

347

348

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

474

IEEE Press

445 Hoes Lane Piscataway, NJ 08854

IEEE Press Editorial Board

Sarah Spurgeon,

Editor in Chief

Jón Atli Benediktsson

Andreas Molisch

Diomidis Spinellis

Anjan Bose

Saeid Nahavandi

Ahmet Murat Tekalp

Adam Drobot

Jeffrey Reed

Peter (Yong) Lia

Thomas Robertazzi

Machine Learning Algorithms for Signal and Image Processing

Edited by

Deepika GhaiLovely Professional University, IN

Suman Lata TripathiLovely Professional University, IN

Sobhit SaxenaLovely Professional University, IN

Manash ChandaMeghnad Saha Institute of Technology, IN

Mamoun AlazabCharles Darwin University, AS

Copyright © 2023 by The Institute of Electrical and Electronics Engineers, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per‐copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750‐8400, fax (978) 750‐4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748‐6011, fax (201) 748‐6008, or online at http://www.wiley.com/go/permission.

Trademarks: Wiley and the Wiley logo are trademarks or registered trademarks of John Wiley & Sons, Inc. and/or its affiliates in the United States and other countries and may not be used without written permission. All other trademarks are the property of their respective owners. John Wiley & Sons, Inc. is not associated with any product or vendor mentioned in this book.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor the author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762‐2974, outside the United States at (317) 572‐3993 or fax (317) 572‐4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our website at www.wiley.com.

Library of Congress Cataloging‐in‐Publication Data Applied for:

Hardback ISBN: 9781119861829

Cover Design: WileyCover Image: © Ryzhi/Shutterstock

Editor Biography

Dr. Deepika Ghai received her Ph.D in the area of signal and image processing from Punjab Engineering College, Chandigarh. She received her M.Tech in VLSI Design & CAD from Thapar University, Patiala, and B.Tech in electronics and communications engineering from Rayat Institute of Engineering and Technology, Ropar. She is an Assistant Professor at Lovely Professional University with more than 8 years' academic experience. She received the Dr. C.B. Gupta Award in 2021 at Lovely Professional University. She has published more than 30 research papers in refereed journals and conferences. She has worked as a session chair, conference steering committee member, editorial board member, and reviewer in international/national IEEE journals and conferences. She has also published edited book “Health Informatics and Technological Solutions for Coronavirus ()” in CRC Taylor & Francis. She is associated as a life member of the Indian Science Congress. Her area of expertise includes signal and image processing, biomedical signal and image processing, and VLSI signal processing.

Dr. Suman Lata Tripathi received her Ph.D. in the area of microelectronics and VLSI from MNNIT, Allahabad. She received her M.Tech in electronics engineering from UP Technical University, Lucknow, and B.Tech in electrical engineering from Purvanchal University, Jaunpur. In 2022 she has worked as are mote post‐doc researcher at Nottingham Trent University, London, UK. She is a Professor at Lovely Professional University and has more than 19 years' academic experience. She has published more than 72 research papers in refereed IEEE, Springer, Elsevier, and IOP science journals and conferences. She has also been awarded 13 Indian patents and 2 copyrights. She has organized several workshops, summer internships, and expert lectures for students. She has worked as a session chair, conference steering committee member, editorial board member, and peer reviewer in international/national IEEE, Springer, Wiley journals and conferences, etc. She received the “Research Excellence Award” in 2019 and “Research Appreciation Award” in 2020, 2021 at Lovely Professional University, India. She received the best paper at IEEE ICICS‐2018. She has edited and authored more than 15 books in different areas of electronics and electrical engineering. She has edited works for Elsevier, CRC Taylor and Francis, Wiley‐IEEE Press, Nova Science, Apple Academic Press, etc. She is also working as a book series editor for “Smart Engineering Systems” and a conference series editor for “Conference Proceedings Series on Intelligent Systems for Engineering Designs” with CRC Press. She is the guest editor of a special issue in “Current Medical Imaging” Bentham Science. She is a senior member, IEEE, fellow at IETE, and life member at ISC and is continuously involved in different professional activities along with academic work. Her area of expertise includes microelectronics device modeling and characterization, low power VLSI circuit design, VLSI design of testing, and advanced FET design for IoT, embedded system design, reconfigurable architecture with FPGAs, and biomedical applications.

Dr. Sobhit Saxena received his Ph.D. from IIT Roorkee in the area of nanotechnology. He did his M.Tech in VLSI and B.E. in electronics and communication engineering. His area of expertise includes nanomaterial synthesis and characterization, electrochemical analysis and modeling, and simulation of CNT‐based interconnects for VLSI circuits. He has designed a new hybrid system of Li‐ion batteries and supercapacitors for energy storage applications. He worked as a (scanning electron microscopy) operator for four years against MHRD fellowship. He has a vast teaching experience of more than 14 years in various colleges and universities. Currently, he is working as an Associate Professor in the School of Electronics and Electrical Engineering, Lovely Professional University. He has been awarded the “Perfect Award” four times in consecutive years for achieving 100% result. He has published more than 10 research papers in SCI/Scopus indexed journals and about 20 papers in reputed international conferences/non‐indexed journals. He has filed three patents, published an edited book “Advanced VLSI Design and Testability Issues” with CRC Press, and two book chapters. He has also published one authored book, Digital VLSI Design and Simulation with Verilog, with Wiley. He is an IEEE member and a reviewer at various refereed SCI/Scopus indexed journals and conference proceedings. He also has industrial exposure in two different companies related to manufacturing (PCB) and broadband communication.

Dr. Manash Chanda graduated in electronics and communication engineering from Kalyani Govt. Engineering College in 2005. He obtained his M.Tech degree in VLSI and microelectronics from Jadavpur University. He completed his Ph.D in engineering from ETCE Dept., Jadavpur University, in 2018. At present, he is working as an Assistant Professor in the Department of ECE, Meghnad Saha Institute of Technology, since February 2006. He is a member of IEEE and is currently a member of IEEE Electron Device Society and Solid State Circuit Society. Dr. Chanda is the co‐founder of IEEE Student Branch and ED MSIT Student Branch Chapter. At present, he is the Chapter Advisor of ED Meghnad Saha Institute of Technology Student Branch Chapter. Also, he is the Vice Chairman of ED Kolkata Chapter. He served as the Secretary of IEEE ED MSIT SBC from January 2018 to December 2019. He has published more than 65 refereed research papers and conference proceedings. His current research interest spans around the study of analytical modeling of sub 100‐nm MOSFETs and nanodevices considering quantum mechanical effects, low‐power VLSI designs, SPICE modeling of nanoscale devices, memory designs, etc. He has published papers in refereed international journals of reputed publishers like IEEE, Elsevier, IET, Springer, Wiley, to name a few. He is the reviewer of many reputed international journals and conferences like IEEE TCAS, IEEE TVLSI, IEEE TED, Solid State Circuits (Elsevier), Journal of Computational Electronics (Springer), International Journal of Numerical Modeling: Electronic Networks, Devices and Fields (Wiley), International Journal of Electronics (Taylor and Francis), etc. He is the recipient of University Gold medal in M. Tech from Jadavpur University in 2008. One of his projects was selected in the Top 10 VLSI project design category (including B. Tech and M.Tech) all over INDIA, organized by CADENCE DESIGN CONTEST, BANGALORE, India in 2010.

Dr. Mamoun Alazab is an associate professor at the College of Engineering, IT, and Environment, and the Inaugural Director of the NT Academic Centre for Cyber Security and Innovation (ACCI) at Charles Darwin University, Australia. He is a cyber‐security researcher and practitioner with industry and academic experience. His research is multidisciplinary and focuses on cyber security and digital forensics of computer systems with a focus on cybercrime detection and prevention. He has published more than 300 research papers (>90% in Q1 and in the top 10% of journal articles, and more than 100 in IEEE/ACM Transactions) and 15 authored/edited books. He received several awards including the NT Young Tall Poppy (2021) from the Australian Institute of Policy and Science (AIPS), IEEE Outstanding Leadership Award (2020), the CDU College of Engineering, IT and Environment Exceptional Researcher Award in (2020) and (2021), and 4 Best Research Paper Awards. He is ranked in top 2% of world's scientists in the subfield discipline of Artificial Intelligence (AI) and Networking & Telecommunications (Stanford University). He was ranked in the top 10% of 30k cyber security authors of all time. Professor Alazab was named in the 2022 Clarivate Analytics Web of Science list of Highly Cited Researchers, which recognizes him as one of the world's most influential researchers of the past decade through the publication of multiple highly cited papers that rank in the top 1% by citations for field and year in Web of Science. He delivered more than 120 keynote speeches, chaired 56 national events and more than 90 international events; on program committees for 200 conferences. He serves as the Associate Editor of IEEE Transactions on Computational Social Systems, IEEE Transactions on Network and Service Management (TNSM), ACM Digital Threats: Research and Practice, Complex & Intelligent Systems.

List of Contributors

Souid Abdelbaki

Department of Electrical Engineering

MACS Research Laboratory RL16ES22

National Engineering School of Gabes

Gabes University

Gabes

Tunisia

Faris Almalki

Department of Computer Engineering

College of Computers and Information Technology

Taif University

Taif

Kingdom of Saudi Arabia

J. Anil Raj

Department of Electronics and Communication

Muthoot Institute of Technology and Science

Kochi

Kerala

India

and

Department of Computer Science

Cochin University of Science and technology

Kochi

Kerala

India

Sundaram Arun

Department of Electronics and Communication Engineering

Jerusalem College of Engineering

Chennai

India

Anterpreet K. Bedi

Department of Electrical and Instrumentation Engineering

Thapar institute of Engineering and Technology

Patiala

Punjab

India

Suman Bera

Department of Computer Science and Engineering

National Institute of Technology

Silchar

Assam

India

Hutashan V. Bhagat

Department of Computer Science and Engineering

Sant Longowal Institute of Engineering and Technology

Longowal

Sangrur

India

Anupam Biswas

Department of Computer Science and Engineering

National Institute of Technology

Silchar

Assam

India

Angshuman Bora

Department of Computer Science and Engineering

National Institute of Technology

Silchar

Assam

India

Yogini D. Borole

Department of E & TC Engineering

G H Raisoni College of Engineering and Management

SPPU Pune University

Pune

India

Subham Chakraborty

Department of Computer Science and Engineering

National Institute of Technology

Silchar

Assam

India

Koneti Chandra Sekhar

School of Electronics and Electrical Engineering

Lovely Professional University

Phagwara

Punjab

India

Pratik Chattopadhyay

Department of Computer Science and Engineering

Indian Institute of Technology (Banaras Hindu University)

Varanasi

India

Palungbam R. Chanu

Electronics and Communication Engineering

NIT Nagaland

Chumukedima

Nagaland

Kaustav Chaudhury

Electronics and Communication Engineering

Heritage Institute of Technology

Anandapur

Kolkata

India

Aneeta Christopher

Department of Electronics and Communication Engineering

National Institute of Technology Calicut

Calicut

Kerala

India

Debangshu Dey

Department of Electrical Engineering

Jadavpur University

Kolkata

West Bengal

India

Thangaraju Dineshkumar

Department of Electronics and Communication Engineering

Kongunadu College of Engineering and Technology

Trichy

India

Paul Fieguth

Vision Image Processing Lab

Department of Systems Design Engineering

University of Waterloo

Waterloo

Canada

Biswarup Ganguly

Department of Electrical Engineering

Meghnad Saha Institute of Technology

Maulana Abul Kalam Azad University of Technology

West Bengal

India

Deepika Ghai

School of Electronics and Electrical Engineering

Lovely Professional University

Phagwara

Punjab

India

R. Hari Kishan

Department of Electronics and Communication Engineering

National Institute of Technology Calicut

Calicut

Kerala

India

Samrina Hussain

Department of Drug Design and Pharmacology

University of Copenhagen

Denmark

Sumam M. Idicula

Department of Artificial Intelligence and Data Science

Muthoot Institute of Technology and Science

Kochi

Kerala

India

Hemlata M. Jadhav

Electronics and Telecommunication Department

Marathwada Mitra Mandal's College of Engineering

Pune

India

Makarand M. Jadhav

Electronics and Telecommunication Department

NBN Sinhgad School of Engineering

Pune

India

Neelu Jain

Electronics and Communication Engineering Department

Punjab Engineering College (Deemed to be University)

Chandigarh

India

Smita Kaloni

Department of Civil Engineering

National Institute of Technology

Uttarakhand

India

Maheshkumar H. Kolekar

Department of Electrical Engineering

Indian Institute of Technology

Patna

Bihar

India

Ranganathan Krishnamoorthy

Centre for nonlinear Systems

Chennai Institute of Technology

Chennai

India

Ashish Kumar

Department of Computer Science and Engineering

Indian Institute of Technology (Banaras Hindu University)

Varanasi

India

Kanak Kumar

Electronics Engineering Department

IEEE Member, Indian Institute of Technology (Banaras Hindu University)

Varanasi

India

Sachin Kumar

Department of Instrumentation and Control Engineering

Dr B R Ambedkar National Institute of Technology

Jalandhar

India

Sandeep Kumar

Department of Electronics and Communications

Sreyas Institute of Engineering and Technology

Hyderabad

Telangana

India

Sanjeev Kumar

Department of BioMedical Applications (BMA)

Central Scientific Instruments Organisation (CSIO)‐CSIR

Chandigarh

India

P. Leninpugalhanthi

Department of EEE

Sri Krishna College of Technology

Coimbatore

Tamil Nadu

India

Swanirbhar Majumder

Department of Information Technology

Tripura University

Agartala

Tripura

India

Debashish Malakar

Department of Computer Science and Engineering

National Institute of Technology

Silchar

Assam

India

Mudasir Maqbool

Department of Pharmaceutical Sciences

University of Kashmir

Hazratbal

Srinagar

India

Madhusudhan Mishra

Department of ECE

NERIST

Nirjuli

Arunachal Pradesh

India

Indiran Mohan

Department of Computer science and Engineering

Prathyusha Engineering College

Chennai

India

Altaf Mulani

Electronics and Telecommunication Department

SKNSCOE

Pandharpur

India

Sugata Munshi

Department of Electrical Engineering

Jadavpur University

Kolkata

West Bengal

India

Swarup Nandi

Department of Information Technology

Tripura University

Agartala

Tripura

India

Mohamed A. Naiel

Vision Image Processing Lab

Department of Systems Design Engineering

University of Waterloo

Waterloo

Canada

Gadamsetti Narasimha Deva

School of Electronics and Electrical Engineering

Lovely Professional University

Phagwara

Punjab

India

Soufiene B. Othman

Department of Telecom, PRINCE Laboratory Research, IsitCom, Hammam Sousse, Higher Institute of Computer Science and Communication Techniques

University of Sousse

Sousse

Tunisia

Shanmugam Padmapriya

Department of Computer Science Engineering

Loyola Institute of Technology

Chennai

India

P. Pandiyan

Department of EEE

KPR Institute of Engineering and Technology

Coimbatore

Tamil Nadu

India

Pooja

Department of Instrumentation and Control Engineering

Dr B R Ambedkar National Institute of Technology

Jalandhar, India

Dasari L. Prasanna

Department of Electronics and Communication Engineering

Lovely Professional University

Phagwara

Punjab

India

Vuyyuru Prashanth

School of Electronics and Electrical Engineering

Lovely Professional University

Phagwara

Punjab

India

Himanshu Priyadarshi

Department of Electrical Engineering

Manipal University Jaipur

Jaipur

India

Ranjeet P. Rai

Department of Electronics Engineering

National Institute of Technology

Uttarakhand

India

Zobeir Raisi

Vision Image Processing Lab

Department of Systems Design Engineering

University of Waterloo

Waterloo

Canada

Husam Rajab

Department of Telecommunications and Media Informatics

Budapest University of Technology and Economics

Budapest

Hungary

Ramya

Department of Electronics and Communication Engineering

Sri Ramakrishna Engineering College

Anna University (Autonomous)

Coimbatore

India

Roshani Raut

Department of Information Technology

Pimpri Chinchwad College of Engineering

Pune

India

Abhisek Ray

Department of Electrical Engineering

Indian Institute of Technology

Patna

Bihar

India

Rehab A. Rayan

Department of Epidemiology

High Institute of Public Health

Alexandria University

Alexandria

Egypt

Thummala Reddychakradhar Goud

School of Electronics and Electrical Engineering

Lovely Professional University

Phagwara

Punjab

India

Hedi Sakli

Department of Electrical Engineering

MACS Research Laboratory RL16ES22

National Engineering School of Gabes

Gabes University

Gabes

Tunisia

and

EITA Consulting 5 Rue du Chant des oiseaux

Montesson

France

S. Saravanan

Department of EEE

Sri Krishna College of Technology

Coimbatore

Tamil Nadu

India

R. Senthil Kumar

Department of EEE

Sri Krishna College of Technology

Coimbatore

Tamil Nadu

India

Shagun Sharma

Department of Electronics Engineering

National Institute of Technology

Uttarakhand

India

Ashish Shrivastava

Faculty of Engineering and Technology

Shri Vishwakarma Skill University

Gurgaon

India

Ghanapriya Singh

Department of Electronics Engineering

National Institute of Technology

Uttarakhand

India

Kulwant Singh

Department of Electronics and Communication Engineering

Manipal University Jaipur

Jaipur

India

Manminder Singh

Department of Computer Science and Engineering

Sant Longowal Institute of Engineering and Technology

Longowal

Sangrur

India

Yeshwant Singh

Department of Computer Science and Engineering

National Institute of Technology

Silchar

Assam

India

Siva Sakthi

Department of Biomedical Engineering

Sri Ramakrishna Engineering College

Anna University (Autonomous)

Coimbatore

India

Gannamani Sriram

School of Electronics and Electrical Engineering

Lovely Professional University

Phagwara

Punjab

India

Sachin Srivastava

Department of Computer Science and Engineering

Indian Institute of Technology (Banaras Hindu University)

Varanasi

India

P.V. Sudeep

Department of Electronics and Communication Engineering

National Institute of Technology Calicut

Calicut

Kerala

India

Ramesh K. Sunkaria

Department of Electronics & Communication Engineering

Dr B R Ambedkar National Institute of Technology

Jalandhar

Punjab

India

K.P. Suresh

Department of EEE

Sri Krishna College of Technology

Coimbatore

Tamil Nadu

India

Ranganathan Thiagarajan

Department of Information Technology

Prathyusha Engineering College

Chennai

India

Suman Lata Tripathi

School of Electronics & Electrical Engineering

Lovely Professional University

Phagwara

Punjab

India

Karan Veer

Department of Instrumentation and Control Engineering

Dr B R Ambedkar National Institute of Technology

Jalandhar

India

Sidhant Yadav

Department of Electronics Engineering

National Institute of Technology

Uttarakhand

India

Georges Younes

Vision Image Processing Lab

Department of Systems Design Engineering

University of Waterloo

Waterloo

Canada

Imran Zafar

Department of Bioinformatics and Computational Biology

Virtual University of Pakistan

Lahore

Punjab

Pakistan

John Zelek

Vision Image Processing Lab

Department of Systems Design Engineering

University of Waterloo

Waterloo

Canada

Muhammad Asim M. Zubair

Department of Pharmaceutics

The Islamia University of Bahawalpur

Pakistan

Preface

Machine learning (ML) algorithms for signal and image processing aid the reader in designing and developing real‐world applications to answer societal and industrial needs using advances in ML to aid and enhance speech signal processing, image processing, computer vision, biomedical signal processing, text processing, etc. It includes signal processing techniques applied for pre‐processing, feature extraction, source separation, or data decompositions to achieve ML tasks. It will advance the current understanding of various ML and deep learning (DL) techniques in terms of their ability to improve upon the existing solutions with accuracy, precision rate, recall rate, processing time, or otherwise. What is most important is that it aims to bridge the gap among the closely related fields of information processing, including ML, DL, digital signal processing (DSP), statistics, kernel theory, and others. It also aims to bridge the gap between academicians, researchers, and industries to provide new technological solutions for healthcare, speech recognition, object detection and classification, etc. It will improve upon the current understanding about data collection and data pre‐processing of signals and images for various applications, implementation of suitable ML and DL techniques for a variety of signals and images, as well as possible collaboration to ensure successful design according to industry standards by working in a team. It will be helpful for researchers and designers to find out key parameters for future work in this area. The researchers working on ML and DL techniques can correlate their work with real‐life applications of smart sign language recognition system, healthcare, smart blind reader system, text‐to‐image generation, or vice versa.

The book will be of interest to beginners working in the field of ML and DL used for signal and image analysis, interdisciplinary in its nature. Written by well‐qualified authors, with work contributed by a team of experts within the field, the work covers a wide range of important topics as follows:

Speech recognition, image reconstruction, object detection and classification, and speech and text processing.

Healthcare monitoring, biomedical systems, and green energy.

Real applications and examples, including a smart text reader system for blind people, a smart sign language recognition system for deaf people, handwritten script recognition, real‐time music transcription, smart agriculture, structural damage prediction from earthquakes, and skin lesion classification from dermoscopic images.

How various ML and DL techniques can improve the accuracy, precision rate recall rate, and processing time.

This easy‐to‐understand yet incredibly thorough reference work will be invaluable to professionals in the field of signal and image processing who want to improve their work. It is also a valuable resource for students and researchers in related fields who want to learn more about the history and recent developments in this field.

Acknowledgments

All editors would like to thank the School of Electronics and Electrical Engineering, Lovely Professional University, Phagwara, India; Department of Electronics and Communication Engineering, Meghnad Saha Institute of Technology, Kolkata; and College of Engineering, IT and Environment at Charles Darwin University, Australia; the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF‐2021S1A5A2A03064391); for providing necessary support for completing this book. The authors would also like to thank the researchers who have contributed their chapters to this book.

Section IMachine Learning and Deep Learning Techniques for Image Processing

1Image Features in Machine Learning

Anterpreet K. Bedig1 and Ramesh K. Sunkaria2

1 Department of Electrical and Instrumentation Engineering, Thapar institute of Engineering and Technology, Patiala, Punjab,, India

2 Department of Electronics & Communication Engineering, Dr B R Ambedkar National Institute of Technology, Jalandhar, Punjab,, India

1.1 Introduction

With the speedily inflating visual data accessible in digital format, generating a highly efficient system for different purposes, such as browsing, searching, retrieving, and managing image data from large amount of database has become the high need of the hour. This has made it important for any image‐processing application to take into consideration any complex information embedded in the data. An image may consist of both visual as well as semantic content in it. Visual data can be categorized into general and domain‐specific. On the other hand, semantic form of data can be extracted using either simple textual annotation or by complicated inference processes depending on certain visual subjects. An image consists of thousands of pixels, each having its individual pixel intensity in several color channels. Images can be characterized by the understanding correlation and relationships between these pixels, thus helping draw a separate identity for each image. In general, it is difficult to integrate the entire information in a reasonable running time for image‐processing purposes. Processing an image based on a particular image property like color or image in database may be a tedious task. This is because comparing every image in the database, including all its pixels, gets very difficult. Also, storing the whole image while creating the database might result in a reduction of storage capacity with the user. Hence, in order to overcome these problems, image features are extracted as a representative of each image.

A feature is described as a piece of data that is used to deliver information regarding the content of an image, typically regarding if a particular region of the image consists of some specific characteristics. A feature is any set of information that can be relevant to solving any task related to a particular application. Basically, image features are certain important and salient points on the image that are meaningful and detectable. They consist of relevant information needed to process the images. Features are limited in number and are not affected by irrelevant variations in the input. In cases where input image data is too extensive to be processed, as well as much of it is believed to be redundant, transforming the same into a minimum number of features is of utmost importance.

Features are easier to handle since they remain unchanged even with various image modifications like rotation, scaling, translation, etc. It is important that features of an image should not change with the image transformations. Also, it is required that features have to be insensitive to lighting conditions and color. Working with features using deep‐learning techniques is more efficient and beneficial for various image‐processing applications. The advantages of features for image‐processing applications are described below:

Rather than storing an entire image, only important and crucial features can be acquired and saved.

It helps in preserving large space for data storage.

It works faster, thus saving large storage and retrieval times.

Using features in deep learning image‐processing applications help in improving its accuracy.

Thus, by using image features as an index, more efficient and accurate indexing systems can be generated. No single feature is entirely suitable for describing and representing any image. For instance, in image retrieval systems, color features are more suitable for retrieving colored images. Texture feature is best suited for describing and retrieving visual patterns, various surface properties, etc. In fact, various texture features are used in describing and retrieving medical images. Other than these, shape features are most suitable in describing certain specific shapes and boundaries of real‐world objects and edges. Thus, no single feature can describe or represent an image completely.

This chapter is arranged as follows: Section 1.2 gives an introduction to feature vectors, followed by a detailed study on various low‐level features in Section 1.3. The chapter is concluded in Section 1.4.

1.2 Feature Vector

Image‐processing techniques based on deep learning make use of corresponding visual data that can be described in terms of features [1, 2]. Any measurements in an image that are extractable and are able to characterize it are considered as features [3]. Making use of images' contents in order to index the image set is called feature extraction. For computation, each feature that is extracted is further encrypted in multi‐dimensional vector, known as feature vector. Feature extraction using deep learning helps in reducing the dimensionality by reducing the raw set of data into more manageable groups for processing. The components of the feature vectors are processed by image‐analysis systems, which are further used for comparison among images. Feature vectors for the entire set of images from the database are combined together to result into a feature database.

Feature extraction is a crucial step in construction of the feature characteristics of an image and targets at extraction of pertinent information that can help in any image‐processing application. A good set of features consists of unique information, which can distinguish one object from another. Features need to be as robust as possible so as to refrain from generation of different feature sets for a single object. Features can be categorized into global and local. In case of global features, visible features from the complete image are taken in one go, whereas, in case of local features, the image content‐ID described using visual descriptors from a region or objects from the image [4