DocumentCode :
441898
Title :
The automation recognizable method of toe shape based on fuzzy comprehensive evaluation
Author :
Yang, Shu ; Gao, Li-Qun ; Wang, Rong ; Bian, Li-Ying
Author_Institution :
Inst. of Inf. Sci. & Eng., Northeastern Univ., Shenyang, China
Volume :
5
fYear :
2005
fDate :
18-21 Aug. 2005
Firstpage :
2614
Abstract :
In the paper, one toe shape recognition method based on fuzzy comprehensive evaluation model is proposed according to the fuzzy character of toe image and the transitional character in different shapes. According to the statistic law of the characteristic value of toe shape, the membership functions belonging to different toe shape are constructed respectively from two effective indexes (angular characteristic value and length characteristic value), then this membership degree is taken as the singular evaluation factor, the toe shape is recognized by using comprehensive evaluation model and the rule of maximum membership degree. Using this fuzzy recognizable method can reflect objectively and accurately the characteristic different toe shape. The experimental results show that the fuzzy recognizable method of toe shape is feasible and effective.
Keywords :
fuzzy set theory; pattern recognition; statistical analysis; angular characteristic value; fuzzy character; fuzzy comprehensive evaluation; length characteristic value; maximum membership degree; membership function; scatter degree; statistic law; toe shape recognition; transitional character; Artificial intelligence; Automation; Character recognition; Competitive intelligence; Electronic mail; Image recognition; Paper technology; Pattern recognition; Shape; Statistics; Toe shape; fuzzy comprehensive evaluation; fuzzy recognition; scatter degree technique; the principle of maximum degree;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on
Conference_Location :
Guangzhou, China
Print_ISBN :
0-7803-9091-1
Type :
conf
DOI :
10.1109/ICMLC.2005.1527385
Filename :
1527385
Link To Document :
بازگشت