BROWSE
Detail
Conference Paper
Face Recognition Based on the Selection Method of Facial Valid Vector in DCT Domain
Cited - time in 
Share 
Share
- Authors
- Yun Su Chung, Ki Hyun Kim, Jean Choi, Jang Hee Yoo
- Issue Date
- 2006-07
- Citation
- International Technical Conference on Circuits/Systems, Computers and Communications (ITC-CSCC) 2006, pp.1-4
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- A novel face recognition method based on the selection of facial valid vector in DCT Domain is described in this paper. The method first generates the facial energy map that depicts the energy distribution characteristics of a face image by using energy probability of each DCT coefficient. And then feature vector is extracted through LDA analysis on the DCT coefficients which are overlapped with the energy map. As the method utilizes the facial energy map, it has the merit that it can consider the energy distribution characteristics of face images, which basically show larger energy congestion in horizontal frequency coefficients than in vertical or diagonal frequency coefficients. Experimental results show that the method is so effective to extract facial feature in DCT domain.
- KSP Keywords
- DCT coefficients, DCT domain, Energy congestion, Energy distribution characteristics, Energy map, Energy probability, Face Image, Feature Vector, Recognition method, Selection method, face recognition