Title :
Fuzzy and ISODATA classification of face contours
Author :
Gu, W.A. ; Su, Guang-Da ; Du, Cheng
Author_Institution :
Dept. of Electron. Eng., Tsinghua Univ., Beijing, China
Abstract :
A method derived from fuzzy and ISODATA clustering algorithm is proposed to classify face contours. An improved ASM method is used to get face contours as one of face shape features. By using the modified ISODATA method based on Hausdorff distance, which is more suitable to classify "shape" features, face contours are clustering into 7 classes. Then the fuzzy c-mean clustering method is used to fuzzy categorize the face contours into different classes. Experiment shows that this clustering approach could automatic classify the human face contours fast and reasonably. Moreover, it could help to accelerate the speed of face recognition and improve the accuracy of face recognition.
Keywords :
face recognition; pattern clustering; ISODATA clustering algorithm; active shape model; face contours classification; face recognition; fuzzy c-mean clustering method; Acceleration; Active shape model; Clustering algorithms; Clustering methods; Data mining; Face detection; Face recognition; Feature extraction; Humans; Spatial databases;
Conference_Titel :
Machine Learning and Cybernetics, 2004. Proceedings of 2004 International Conference on
Print_ISBN :
0-7803-8403-2
DOI :
10.1109/ICMLC.2004.1380408
