Advanced Methods for Human Biometrics

Description

Preface Contents Part I Authentication Based on Measurements of Human Characteristics 1 Efficient Fingerprint Analysis Based on Sweat Pore Map 1.1 Introduction 1.2 Related Works 1.3 Proposed Approach 1.3.1 Step 1: Pores Detection 1.3.2 Step 2: Features Extraction 1.3.3 Step 3: Pores Alignment 1.3.4 Step 4: Pores Matching 1.4 Experiments and Performance Evaluation 1.4.1 Data Base 1.4.2 Training and Test Process 1.4.3 Feature Matching 1.4.4 Performance Evaluation 1.5 Conclusion References 2 Fingerprint Recognition Based on Level Three Features 2.1 Introduction 2.2 Biometry Background 2.2.1 Biometric Systems 2.2.2 Biology of the Fingerprint 2.3 Pores Detection 2.3.1 Related Works 2.3.2 Proposed Method 2.4 Pores Matching 2.4.1 Related Works 2.4.2 Proposed Method 2.5 Experimental Results 2.5.1 Database 2.5.2 Pores Detection 2.5.3 Recognition 2.6 Conclusion References 3 Fractal Analysis for Iris Multimodal Biometry 3.1 Introduction 3.2 Related Works 3.3 Feature Extraction Based on Fractal Analysis 3.4 Uni-Modal Recognition System 3.4.1 PBMLTiris Database Description 3.4.2 Pre-processing 3.4.3 Iris Segmentation (Daugman's Operator) 3.4.4 Normalization Based on the Pseudo-Polar Method (Masšek, ch3AmenispsbibspsMaek2003RecognitionOH) 3.4.5 Matching 3.5 Multi-modal Recognition System 3.5.1 Limitations of Uni-Modal Recognition System (Singh et al., ch3Amenispsbibspssingh2019comprehensive) 3.5.2 Fusion Sources 3.5.3 Fusion Levels 3.6 Experimental Results 3.6.1 Segmentation Results 3.6.2 Uni-Modal System Evaluation 3.6.3 Feature Level Fusion Results 3.6.4 Sensor Level Fusion Results 3.6.5 Score Level Fusion Results 3.7 Discussion and Conclusion References Part II Authentication by Biological Signals 4 Security with ECG Biometrics 4.1 Biometrics Definition 4.2 Biometrics with ECG 4.3 ECG Biometrics Approaches 4.3.1 Fiducial Approaches 4.3.2 Non-fiducial Approaches 4.4 ECG Signal Filters 4.5 ECG Biometric Classifiers 4.6 Evaluation of ECG Biometrics 4.7 Conclusion References 5 ECG Biometric System for Human Recognition Based on the Possibility Theory 5.1 Introduction 5.2 Possibility Theory 5.2.1 Possibility Distribution 5.2.2 Transformation from Probability Distribution to Possibility Distribution 5.3 Methodology 5.3.1 ECG Signal Pre-processing 5.3.2 Feature Extraction 5.3.3 Possibility Theory Based ECG Classification 5.3.4 Experimental Results and Discussion 5.4 Conclusion References 6 Surface EMG Based Biometric Person Authentication by a Grasshopper Optimized SVM Algorithm 6.1 Introduction 6.2 Biometry Based on sEMG Signals 6.3 Hybrid Grasshopper Optimization Algorithm and Support Vector Machine (GOA-SVM) 6.3.1 Grasshopper Optimization Algorithm (GOA) 6.3.2 GOA-SVM 6.4 Experimental Results 6.5 Conclusion References Part III Algorithm Based Methods of Multimodal Authentication 7 Tracklet and Signature Representation Using Part Appearance Mixture Approach in the Context of Multi-shot Person Re-Identification 7.1 Introduction 7.2 Main Challenges of Person Re-ID 7.3 Related Works 7.4 Person Re-ID Process 7.4.1 Detection 7.4.2 Multi-object Tracking 7.5 Part Appearance Mixture (PAM) Approach 7.5.1 Signature Representation 7.5.2 Similarity Metric for Signature Representation 7.5.3 Distance Computation Between Signatures 7.6 Experiments and Results 7.6.1 Datasets 7.6.2 Performance Evaluation 7.6.3 Evaluation of Signature Representation Quality 7.7 Conclusion References 8 A Novel Approach for Speaker Recognition in Degraded Conditions 8.1 Introduction 8.2 Related Works 8.3 Proposed Approach 8.3.1 Pre-processing 8.3.2 Feature Extraction 8.3.3 Classification 8.4 Experimental Results 8.5 Conclusion References 9 Visual Methods for Sign Language Recognition: A Modality-Based Review 9.1 Introduction 9.2 Human Actions Recognition Pipeline 9.3 Unimodal Methods 9.3.1 Recognition from Joint Streams 9.3.2 Recognition from RGB Streams 9.3.3 Recognition from Depth Streams 9.3.4 Unimodal Temporal Segmentation Approaches 9.4 Multi-modal Methods 9.4.1 Multi-modal Datasets for HAR 9.4.2 Multi-modal Fusion Approaches 9.4.3 Multi-modal Datasets for 3D FEs Recognition 9.4.4 Multi-modal Approaches for 3D FEs Recognition 9.5 Main Contributions Related to SL Recognition 9.5.1 SL Datasets 9.5.2 SL Visual-Recognition Based Works 9.6 Conclusion and Discussion 9.6.1 Datasets Level 9.6.2 Approaches Level 9.6.3 Commercial Solutions Level References 10 A Software Architecture for Developing Disease Registries 10.1 Introduction 10.2 Related Work 10.2.1 Technology 10.2.2 Data 10.2.3 Knowledge 10.2.4 Analytics 10.2.5 Services 10.2.6 Security 10.2.7 Sharing 10.3 Proposed Software Architecture 10.3.1 Technology Layer 10.3.2 Data Layer 10.3.3 Knowledge Layer 10.3.4 Analytics Layer 10.3.5 Service Layer 10.3.6 Security and Privacy 10.3.7 Sharing 10.3.8 Interactions 10.4 Use Cases 10.5 Conclusion References Part IV Biomedical Characteritics 11 3D Textures Analysis Based on Features Extraction 11.1 Introduction 11.2 Methods of Texture Measures 11.2.1 Decimal Descriptor Patterns (DDP) 11.2.2 Local Binary Patterns 11.2.3 Grey Level Co-occurrence Matrix Method 11.3 Experiments and Results 11.3.1 Databases 11.3.2 Phases of Simulation 11.3.3 3D MR Brain Images Analysis 11.3.4 3D Face Analysis 11.3.5 Discussion 11.4 Conclusion References 12 Image Processing and Analysis for Decision Making Applied to Melanoma 12.1 Introduction 12.2 About Melanoma 12.3 Diagnostic Aid System Based on Score Computation 12.3.1 Images Acquisition 12.3.2 Images Pretreatment 12.3.3 Lesion Detection 12.3.4 Interpretation of Medical Images 12.4 Diagnostic Aid System Based on Machine Learning 12.4.1 Images Acquisition 12.4.2 Pretreatment of Dermatoscopic Images 12.4.3 Segmentation of Lesion Based on Region Growing Method 12.4.4 Skin Lesion Analysis 12.5 Experimental Results and Discussion 12.5.1 Approach Based on the MultiOtsu Principle 12.5.2 Approach Based on the Region Growing Method 12.5.3 Evaluation and Discussion 12.6 Conclusion References 13 Biomedical Computer Aided Design Systems: Application to Alzheimer Disease 13.1 Introduction 13.2 Proposed Methodology 13.3 Previous Works 13.3.1 Partial Least Square (PLS) 13.3.2 Kernel Partial Least Square (KPLS) 13.4 Proposed Downsized KPLS Method (DPLS) 13.5 Optimization with Multi-objective Optimization Algorithm 13.5.1 Principle 13.5.2 Selection of Kernel Parameter with Multi-Objective Optimization Algorithm 13.6 Classification Using Neural Networks 13.7 Experiments 13.7.1 Experiments on ADNI Dataset 13.7.2 Experiments on OASIS Dataset 13.8 Conclusion and Future Work References