Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani
Over a few decades, many computer vision systems haven been developed. One of the applications related to computer vision is face recognition and was being interested by many researches. This project is all about implementing the back-propagation neural network algorithm in classification of face ex...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 語言: | English |
| 出版: |
2005
|
| 主題: | |
| 在線閱讀: | https://ir.uitm.edu.my/id/eprint/9393/1/TD_MAZURAINI%20GHANI%20CS%2005_24.pdf |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| id |
my-uitm-ir.9393 |
|---|---|
| record_format |
uketd_dc |
| spelling |
my-uitm-ir.93932017-01-25T07:09:37Z Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani 2005 Ghani, Mazuraini Neural networks (Computer science) Pattern recognition systems Over a few decades, many computer vision systems haven been developed. One of the applications related to computer vision is face recognition and was being interested by many researches. This project is all about implementing the back-propagation neural network algorithm in classification of face expression. This project has 3 objectives. The first objective is to collect and digitized images with different expressions which is neutral, happy and angry. The second is to design and develop a prototype for classifying human emotions by face expression recognition of a given image, using back propagation neural network. The last is to study and experiment the suitable edge detection techniques for binary facial image. There are two important phases that were focused in developing this project. The phases are pre-processmg phase and neural network design phase. In preprocessing phase, s detail studies and intensive experiments were conducted to obtain a suitable method of segmentation. Meanwhile, in the neural network design and implementation phase, intensive experiments have been conducted to obtained appropriate design and parameter value of neural network. In this project, the suitable method of segmentation is local adaptive threshold. However, the performances of neural network in learning and classification task should be enhanced by redesigning and conducting experiment on other learning algorithm than back-propagation. 2005 Thesis https://ir.uitm.edu.my/id/eprint/9393/ https://ir.uitm.edu.my/id/eprint/9393/1/TD_MAZURAINI%20GHANI%20CS%2005_24.pdf text en public other degree Universiti Teknologi MARA Faculty of Computer and Mathematical Science |
| institution |
Universiti Teknologi MARA |
| collection |
UiTM Institutional Repository |
| language |
English |
| topic |
Neural networks (Computer science) Pattern recognition systems |
| spellingShingle |
Neural networks (Computer science) Pattern recognition systems Ghani, Mazuraini Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani |
| description |
Over a few decades, many computer vision systems haven been developed. One of the applications related to computer vision is face recognition and was being interested by many researches. This project is all about implementing the back-propagation neural network algorithm in classification of face expression. This project has 3 objectives. The first objective is to collect and digitized images with different expressions which is neutral, happy and angry. The second is to design and develop a prototype for classifying human emotions by face expression recognition of a given image, using back propagation neural network. The last is to study and experiment the suitable edge detection techniques for binary facial image. There are two important phases that were focused in developing this project. The phases are pre-processmg phase and neural network design phase. In preprocessing phase, s detail studies and intensive experiments were conducted to obtain a suitable method of segmentation. Meanwhile, in the neural network design and implementation phase, intensive experiments have been conducted to obtained appropriate design and parameter value of neural network. In this project, the suitable method of segmentation is local adaptive threshold. However, the performances of neural network in learning and classification task should be enhanced by redesigning and conducting experiment on other learning algorithm than back-propagation. |
| format |
Thesis |
| qualification_name |
other |
| qualification_level |
Bachelor degree |
| author |
Ghani, Mazuraini |
| author_facet |
Ghani, Mazuraini |
| author_sort |
Ghani, Mazuraini |
| title |
Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani |
| title_short |
Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani |
| title_full |
Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani |
| title_fullStr |
Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani |
| title_full_unstemmed |
Face expression recognition using artificial neural network (ANN) / Mazuraini Ghani |
| title_sort |
face expression recognition using artificial neural network (ann) / mazuraini ghani |
| granting_institution |
Universiti Teknologi MARA |
| granting_department |
Faculty of Computer and Mathematical Science |
| publishDate |
2005 |
| url |
https://ir.uitm.edu.my/id/eprint/9393/1/TD_MAZURAINI%20GHANI%20CS%2005_24.pdf |
| _version_ |
1783733155782459392 |