Acknowledgement | 第6-7页 |
Abstract | 第7页 |
This dissertation is divided into eight Chapters | 第9-10页 |
Abbreviations | 第10-13页 |
Chapter 1 | 第13-17页 |
1-Overview | 第14-17页 |
1.1 Motivation and problem description | 第14-15页 |
1.2 Suggested solution | 第15-17页 |
Chapter 2:Theoretical Background and Previous Work | 第17-54页 |
Part 1: Theory | 第18-44页 |
2.1 Human's muscles | 第18-24页 |
2.1.1 Anatomy of human forearm | 第19-24页 |
2.2 Electromyographic EMG signal | 第24-35页 |
2.2.1 Introduction | 第24-25页 |
2.2.2 EMG-History | 第25-26页 |
2.2.3 EMG-Signal processing: EMG recording-Noise effect | 第26-28页 |
2.2.4 Feature extraction | 第28-32页 |
2.2.5 Wavelet analysis | 第32-35页 |
2.3 Pattern recognition | 第35-44页 |
2.3.1 Neural network-Historical overviews | 第36-37页 |
2.3.2 Supervised and unsupervised learning | 第37-38页 |
2.3.3 Classification methods | 第38-44页 |
Part 2:Myoelectric Pattern Recognition System Using Different ClassificationApproaches:A Review | 第44-54页 |
2.1 Literature review-Myoelectric pattern recognition | 第44-53页 |
2.2 Conclusion | 第53-54页 |
Chapter 3:Recognition of Five Hand Motions Patterns Using RBF Neural NetworkBased on Surface Electromyography Signal | 第54-64页 |
3.1 Introduction | 第55页 |
3.2 Subjects-Material-Experimental protocol | 第55-59页 |
3.2.1 Subjects | 第55-56页 |
3.2.2 Material | 第56-57页 |
3.2.3 Experimental protocol | 第57-59页 |
3.3 Feature selection and classification step | 第59-60页 |
3.4 Result and discussion | 第60-62页 |
3.5 Conclusion,limitaiton and future work | 第62-64页 |
Chapter 4:An Investigation of Myoelectric Pattern Recognition Based on WaveletAnalysis | 第64-75页 |
4.1 Introduction | 第65页 |
4.2 Wavelet analysis in myoelectric pattern recognition system | 第65-67页 |
4.3 The proposed myoelectric PR system based on wavelet coefficients | 第67-69页 |
4.4 Results and discussion | 第69-73页 |
4.5 Conclusion | 第73页 |
4.6 Limitations and future work | 第73-75页 |
Chapter 5:Pattern Recognition of Eight Hand Motions Using FeatureExtraction of Forearm EMG Signal Based on Statistical Analysis | 第75-95页 |
5.1 Introduction | 第76-77页 |
5.2 Experimental protocol | 第77-81页 |
5.3 Feature weighing the proposed pattern recognition system | 第81-83页 |
5.4 Results and discussions | 第83-94页 |
5.4.1 Part One:Feature weighting | 第83-87页 |
5.4.2 Part Two:Feature combination | 第87-88页 |
5.4.3 Part Three:Select the kernels functions of support vector machine (SVM) | 第88-90页 |
5.4.4 Part Four:Selecting best wavelet family | 第90-93页 |
5.4.5 Part Five:Optimization of K-nearest neighbor (K-NN) algorithm based on k-value | 第93-94页 |
5.5 Conclusion | 第94-95页 |
Chapter 6:Different Dimensionality Reduction Methods for ClassificationEleven Hand Motions Using Wavelet Packet Multilayer Neural Network andArtificial Bee Colony Algorithm | 第95-101页 |
6.1 Introduction | 第96-97页 |
6.2 Multilayer perceptron back propagation neural network | 第97-98页 |
6.3 Result and discussion | 第98-99页 |
6.4 Proposed ABC-MLPNN algorithm | 第99-100页 |
6.5 Conclusion | 第100-101页 |
Chapter 7:A Novel Pattern Recognition Approach Using GA-LibSVM HybridModel for Classification Eleven Hand Motions | 第101-107页 |
7.1 Introduction | 第102-103页 |
7.2 Block diagram of the proposed myoelectric pattern recognition system | 第103-105页 |
7.2.1 Data acquisition | 第103页 |
7.2.2 Dimensionality reduction | 第103页 |
7.2.3 Feature extraction | 第103-104页 |
7.2.4 The proposed hybrid GA-LibSVM pattern recognition model | 第104-105页 |
7.3 Results and discussion | 第105-106页 |
7.4 Conclusion | 第106-107页 |
Chapter 8:Conclusion, Future Work and Limitations | 第107-110页 |
Academic Publications | 第110-111页 |
Reference | 第111-128页 |