Fuzzy C mean clustering in off-line handwriting signature verfication system
This research is the first research that suggests the usage of the stable region of a signature for verification purpose. This research highlighted the design and development of the proposed Fuzzy C Mean clustering to determine the stable segments of a signature generated from windowing segmentation...
Saved in:
主要作者: | |
---|---|
格式: | Thesis |
语言: | English |
出版: |
2006
|
主题: | |
在线阅读: | http://ir.unimas.my/id/eprint/1705/8/2013-02-thLeeBYfull.pdf |
标签: |
添加标签
没有标签, 成为第一个标记此记录!
|
总结: | This research is the first research that suggests the usage of the stable region of a signature for verification purpose. This research highlighted the design and development of the proposed Fuzzy C Mean clustering to determine the stable segments of a signature generated from windowing segmentation process where features could be extracted from signature images. Six features are extracted from each stable segment i.e. Image Size, Ratio of height to width, slant, Maximum horizontal projection, Vertical centre of mass and Horizontal Relative gravity centre/ Horizontal centre of mass. Besides that, this research also demonstrated how to select a best features to represent each determined stable segment. |
---|