Hardware modeling of EMG signal classifier by using discrette wavelet transform and neural network for human computer interaction /
Electromyography (EMG) signal based research is ongoing for the development of simple, robust, user friendly, efficient interfacing devices/systems. An EMG signal based reliable and efficient hand gesture identification system has been developed for human computer interaction, which in turn will inc...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
Kuala Lumpur :
Kulliyyah of Engineering, International Islamic University Malaysia,
2012
|
| Subjects: | |
| Online Access: | Click here to view 1st 24 pages of the thesis. Members can view fulltext at the specified PCs in the library. |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Electromyography (EMG) signal based research is ongoing for the development of simple, robust, user friendly, efficient interfacing devices/systems. An EMG signal based reliable and efficient hand gesture identification system has been developed for human computer interaction, which in turn will increase the quality of life of the disabled or aged people. The acquired and processed EMG signal requires classification before utilizing it in the development of interfacing, which is the most difficult part of the development process. Another difficulty involved in the analysis of the EMG signal is due to its noisy characteristics. Compared to other biosignals, EMG signal contains complicated types of noise that are caused by, for example, inherent equipment noise, electromagnetic radiation, motion artifacts, and the interaction of different tissues. Hence, EMG signal has been denoised first to filter out the unwanted noises by implanting 4-level Daubechies 4-tap (db4) Discrete Wavelet Transform (DWT). Then different predefined hand motions (left, right, up and down) have been detected using Artificial Neural Network (ANN). The capability of learning from examples, the ability to reproduce arbitrary non-linear functions of input, and the highly parallel and regular structure make the ANNs suitable especially for pattern recognition tasks. The EMG pattern signatures for each movement have been extracted from the denoised signals and then the signals have been classified by utilizing the ANN. A back-propagation neural network with Levenberg-Marquardt training algorithm has been used for the detection of gesture. The conventional and most effective time and time-frequency based features (namely MAV, RMS, VAR, SD, ZC, SSC and WL) have been chosen to train the neural network. Experimentally it has been found that Levenberg-Marquardt algorithm based ANN recognizes the desired motions efficiently and takes minimal computation time. The designed ANN has successfully classified the EMG signals obtained from the hand movements and resulting in average success rate is 88.4%, where the best overall performance is achieved 89.2% in a single trial. Afterwards, the designed DWT based denoiser and ANN based classifier have been modeled using Hardware Description Language (HDL) for hardware realization. VHDL (Very High Speed Integrated Circuit Hardware Description Language) has been used to model the algorithm implemented into the target device FPGA (Field Programmable Gate Array). The designed model has been synthesized and fitted into Altera's Stratix III, chipset EP3SE50F780I4L using the Quartus II version 9.1 Web Edition. All the weight and bias values considered in designing the hardware model of ANN have made the design complex, which affects the memory consumption and computational time. No specific way is developed yet to find the weight and bias values which are significant for determining the performance of ANN. Avoiding non-significant weights and biases during hardware modeling may save a lot of memory and computational time. |
|---|---|
| Item Description: | Abstracts in English and Arabic. "A thesis submitted in fulfilment of the requirement for the degree of Master of Science (Electronics Engineering)."--On t.p. |
| Physical Description: | xvii, 157 leaves : ill charts ; 30cm. |
| Bibliography: | Includes bibliographical references (leaves 139-149). |