A guide to the core ideas of human motion capture in a rapidly changing technological landscape
Human Motion Capture and Identification for Assistive Systems Design in Rehabilitation aims to fill a gap in the literature by providing a link between sensing, data analytics, and signal processing through the characterisation of movements of clinical significance. As noted experts on the topic, the authors apply an application-focused approach in offering an essential guide that explores various affordable and readily available technologies for sensing human motion.
The book attempts to offer a fundamental approach to the capture of human bio-kinematic motions for the purpose of uncovering diagnostic and severity assessment parameters of movement disorders. This is achieved through an analysis of the physiological reasoning behind such motions. Comprehensive in scope, the text also covers sensors and data capture and details their translation to different features of movement with clinical significance, thereby linking them in a seamless and cohesive form and introducing a new form of assistive device design literature. This important book:- Offers a fundamental approach to bio-kinematic motions and the physiological reasoning behind such motions- Includes information on sensors and data capture and explores their clinical significance- Links sensors and data capture to parameters of interest to therapists and clinicians- Addresses the need for a comprehensive coverage of human motion capture and identification for the purpose of diagnosis and severity assessment of movement disorders
Written for academics, technologists, therapists, and clinicians focusing on human motion, Human Motion Capture and Identification for Assistive Systems Design in Rehabilitation provides a holistic view for assistive device design, optimizing various parameters of interest to relevant audiences.
Table of Contents
1 Introduction 9
1.1 Human Body - Kinematic Perspective 10
1.2 Musculoskeletal Injuries and Neurological Movement Disorders 11
1.2.1 Musculoskeletal injuries 11
1.2.2 Neuromascular disorders 12
1.3 Sensors in Tele-rehabilitation 15
1.3.1 Opto-electroic Sensing 15
1.3.2 RGB Camera and Microphone 20
1.3.3 Inertial Measurement Unit (IMU) 22
1.4 Model based state estimation and sensor fusion 24
1.4.1 Summary and challenges 24
1.5 Human Motion Encoding in Tele-rehabilitation 25
1.5.1 Human motion encoders in action recognition 25
1.5.2 Human motion encoder in physical tele-rehabilitation 26
1.5.3 Summary and challenge 27
1.6 Patients’ Performance Evaluation 28
1.6.1 Questionnaire based assessment scales 28
1.6.2 Automated kinematic performance assessment 29
1.6.3 Summary and challenge 30
2 Kinematic Performance Evaluation with Non-wearable Sensors 31
2.1 Introduction 32
2.2 Fusion 32
2.2.1 Introduction 32
2.2.2 Linear Model of Human Motion Multi-Kinect System 34
2.2.3 Model-based state estimation 37
2.2.4 Fusion of Information 37
2.2.5 Mitigation of occlusions and optimised positioning 37
2.2.6 Computer simulations and hardware implementation 38
2.3 Encoder 49
2.3.1 Introduction 49
2.3.2 The two-component encoder theory 51
2.3.3 Encoding Methods 52
2.3.4 Dealing with Noise 54
2.3.5 Complex Motion Decomposition using Switching Continuous Hidden Markov Models 56
2.3.6 Canonical Actions and the Action Alphabet 57
2.3.7 Experiments and Results 59
2.4 ADL Kinematic Performance Evaluation 67
2.4.1 Introduction 67
2.4.2 Methodology 68
2.4.3 Experiment Setup 71
2.4.4 Data Analysis and Results 75
2.5 Summary 83
3 BioKinematic Measurement with Wearable Sensors 85
3.1 INTRODUCTION 86
3.2 Kinematic Model 86
3.3 Introduction to Quaternions 86
3.4 Wahba’s Problem 87
3.4.1 Solutions to the Whaba’s Problem 88
3.4.2 Davenport’s q method 89
3.4.3 Quaternion Estimation Algorithm(QUEST) 91
3.4.4 Fast Optimal Attiude Matrix (FOAM) 92
3.4.5 Estimator of the optimal quarternion (ESOQ or ESOQ1) method 92
3.5 Quaternion Propagation 92
3.6 MARG (Magnetic Angular Rates and Gravity) Sensor arrays based Algorithm 93
3.7 Model based estimation of attitude with IMU data 93
3.8 Robust optimisation based approach for Orientation estimation 96
3.9 Implementation of the Orientation estimation 97
3.9.1 Extended Kalman Filter based approach 97
3.9.2 Robust Extended Kalman Filter Implementation 98
3.9.3 Robust Extended Kalman Filter with Linear Measurements 99
3.10 Computer Simulations 99
3.11 Experimental Setup 100
3.12 Results and Discussion 102
3.12.1 Computer Simulations 102
3.12.2 Experiment 104
3.13 Conclusion 105
4 Capturing Finger movements 107
4.1 Introduction 108
4.2 System Overview 111
4.3 Accuracy Improvement of Total Active Movement and Proximal Interphalangeal Joint Angles 111
4.4 Simulation 115
4.5 Trial Procedure 116
4.6 Results 117
4.6.1 Concurrence validity 117
4.6.2 Internal Reliability 119
4.6.3 Time efficiency 119
4.7 Discussions 120
4.8 Approaching finger movement with a new perspective 122
4.9 Reachable Space 124
4.10 Boundary of the Reachable Space 127
4.11 Area of Reachable Space 131
4.12 Experiments 133
4.13 Results and Discussion 134
4.14 Conclusion & Future Work 140
5 Non-contact measurement of respiratory function via Doppler Radar 141
5.1 Introduction 142
5.2 Fundamental Operation of Microwave Doppler Radar 144
5.2.1 Velocity and frequency 144
5.2.2 Correction of I/Q Amplitude and Phase Imbalance 146
5.3 Signal Processing Approach 149
5.3.1 Respiration rate 149
5.3.2 Extracting respiratory signatures 151
5.3.3 Low Pass Filtering (LPF) 153
5.3.4 Discrete Wavelet Transform 154
5.4 Common Data Acquisitions Setup 155
5.5 Capturing the dynamics of respiration 157
5.5.1 Normal Breathing 158
5.5.2 Fast Breathing 158
5.5.3 Slow Inhalation - Fast Exhalation 158
5.5.4 Fast Inhalation - Slow Exhalation 158
5.5.5 Capturing Abnormal Breathing Patterns 160
5.5.6 Breathing Component Decomposition, Analysis and Classification 161
5.6 Capturing Special Breathing Patterns 164
5.6.1 Correlation of Radar Signal with Spirometer in Tidal Volume Estimations 165
5.6.2 Experiment Setup 166
5.6.3 Results 166
5.6.4 Motion Signature from Doppler radar 175
5.6.5 Measurement of Volume In (Inhalation) and Volume Out (Exhalation) 176
5.7 Removal of Motion Artefacts from Doppler Radar based Respiratory Measurements179
5.7.1 Experimental verification 181
5.7.2 Results and Discussion 182
5.7.3 summary 185
5.8 Separation of Doppler Radar based Respiratory Signatures 186
5.8.1 Respiration Sensing Using Doppler Radar 186
5.8.2 Signal Processing -Source Separation (ICA) 187
5.8.3 Experiment Protocol for Real Data Sensing 189
5.8.4 Two Simulated Respiratory Sources 190
5.8.5 Experiment involving real subjects 190
5.8.6 Separation of Hand Motion 193
5.8.7 Conclusion 195
6 Appendix I 203
6.1 Static Estimators 204
6.1.1 Least Squares Estimation 204
6.1.2 Maximum likelihood estimation 204
6.2 Model based estimators 205
6.2.1 Kalman filter (KF) 205
6.3 Particle filter 207
6.3.1 Robust filtering with linear measurements 208
6.3.2 Constrained optimisation 210
