Innovations in Computer Vision and Data Classification: From Pandemic Data Analysis to Environmental and Health Monitoring (EAI/Springer Innovations in Communication and Computing)

Preface
Acknowledgements
Contents
Design and Development of an Integrated Healthcare Platform Using Deep Convolutional Neural Networks (DCNNs)
	1 Background
	2 Methodology
		2.1 Preprocessing
			2.1.1 AlexNet Model
			2.1.2 GoogLeNet Model
			2.1.3 ResNet101 Model
	3 Proposed DCNN Model
		3.1 Input Layer (Layer1)
		3.2 Two-Dimensional Convolutional Layer (Layers 2-11)
		3.3 Batch Normalization Layer (Layers 3, 9, 13, 24, 28 and 31)
		3.4 ReLU Layer
		3.5 Max Pooling Layer (Layers 5, 11, 15, 26 and 33)
		3.6 Dropout Layer
		3.7 Fully Connected Layer (Layers 34-39)
		3.8 Softmax Layer (Layer 39)
		3.9 Output Layers (Layer 40)
	4 Evaluation of Performance Metrics
		4.1 Validation and Training Accuracy
		4.2 Confusion Matrix
		4.3 ROC Curve
	5 Results
	6 Discussion
	7 Conclusion
	8 Activities for Researchers/Practitioners
	References
Advancements in Digital Health Diagnostics: Mathematical Modelling in the Detection of Cancer Cells
	1 Introduction
	2 Background
	3 Methodology
		3.1 Mathematical Formulation
		3.2 Proliferation Area Formulation
		3.3 Quiescent Area Formulation
		3.4 Necrotic Area Formulation
	4 Algorithmic Design
	5 Results
	6 Discussion
	7 Summary
	8 Activities for Researchers/Practitioners
	References
A Case Study on Real-Time Performance Analysis of Maximum Power Point Tracking (MPPT) with Reconfigurable Hardware (FPGA)
	1 Background
	2 PV Ideal Single-Diode Model (ISDM)
		2.1 PV Single-Diode Model (SDM)
		2.2 Maximum Power Point Tracking (MPPT)
	3 Case Study
		3.1 System Architecture
		3.2 MPPT Algorithms
		3.3 Perturb and Observe (P&O)
		3.4 Incremental Conductance (IncCon)
		3.5 DC-DC Boost Converter
		3.6 Design Verification Using MATLAB-Simulink
	4 Field Programmable Gate Array (FPGA)-Based Design
		4.1 Hardware Implementation and Elaborated Design
		4.2 Perturb and Observe (P&O)
		4.3 Incremental Conductance (IncCon)
		4.4 Pulse Width Modulation (PWM)
	5 Design Implementation
		5.1 Software Implementation
		5.2 Hardware Implementation and Testing
	6 PV Module Performance Estimation
	7 Comparison of Speed, Power and Design Scalability
	8 Summary
	9 Activities for Researchers/Practitioners
	References
A System-Level Approach to Sustainable Low Power Sensing: Meeting United Nations Sustainable Development Goals (SDGs)
	1 Background
	2 Proposed Methodology
		2.1 System Design
			2.1.1 Hardware Design
			2.1.2 Firmware Design
			2.1.3 PCB Implementation
	3 Design Analysis
		3.1 Conditions and Results
		3.2 Testing the Load
		3.3 Testing the Solar Charger
		3.4 Testing the DC-DC Boost Converter
	4 Discussion
	5 Conclusion
	6 Activities for Researchers/Practitioners
	References
Fundamentals of Low-Power Neuromorphic Circuit Design with Spiking Neural Networks (SNNs)
	1 Background
	2 Fundamentals of SNN Circuits and Mathematical Models
		2.1 Leaky Integrate and Fire Neuron Model
		2.2 Mathematical Models of Synapses
		2.3 Fundamental Synapse Circuit Design
	3 Methodology for Neuromorphic Circuit Simulation
		3.1 Circuit Analysis and Testing
		3.2 Circuit Description
	4 Circuit Simulations
		4.1 Subthreshold Characteristics
		4.2 Synapses Strength and Time Constant
		4.3 Spike Train Response
	5 Addressing Nonidealities and Optimizing Design
		5.1 First-Order System
		5.2 Biologically Plausible Circuit Model
	6 Discussion
	7 Future Tasks for Students/Researchers
	References
Embodied General AI Require Decision Support System (DSS): An Embedded Design for Electrocardiogram (ECG) Interpretation
	1 Background
	2 Implementation
		2.1 Wavelet Transformation
			2.1.1 Noise Filtration
			2.1.2 Proposed Prototype Wavelet
			2.1.3 Algorithm Description
			2.1.4 Validation
		2.2 Decision Support System (DSS)
	3 Discussion
		3.1 QRS Detection
		3.2 Delineation Results
		3.3 Decision Support System (DSS) Design Results
	4 Discussion
	5 Limitations and Future Work
	6 Summary
	7 Activities for Researchers/Practitioners
	References
Computer Vision-Based Automated Diagnosis for Skin Cancer Detection
	1 Background
	2 Convolutional Neural Network (CNN) Architecture
		2.1 Convolutional Layer
		2.2 Pooling Layer
		2.3 ReLU (Rectified Linear Unit)
		2.4 Fully Connected Layer
		2.5 Loss Layer
	3 Pretrained Convolutional Neural Networks
	4 Proposed Method for Skin Lesion Detection
		4.1 Development of CNNs
		4.2 Dataset Selection
		4.3 Baseline Testing Without Image Processing
		4.4 Imaging Preprocessing to Isolate the Lesions
		4.5 Extended CNN Training
	5 Results
	6 Discussion
	7 Activities for Researchers/Practitioners
	References
Computer Vision-Based Human Activity Detection for Intensive Care Patients: A Case Study Based on Region of Interest (ROI)
	1 Introduction
	2 Implementation Method
		2.1 Proposed Computer Vision Model
		2.2 Data Collection and Preprocessing
		2.3 Image Acquisition System
		2.4 Object Detection
		2.5 ROI Extraction
		2.6 Gradient Comparison
	3 Analysis
	4 Discussion
	5 Summary
	References
Index