Use this resource - and many more! - in your textbook!
AcademicPub holds over eight million pieces of educational content for you to mix-and-match your way.
Two Dimensional Orthogonal Wavelet Features for Image Representation and Recognition
By: Woo, W.L.; Mutelo, R.M.; Dlay, S.S.;
2007 / IEEE / 1-4244-0881-4
This item was taken from the IEEE Conference ' Two Dimensional Orthogonal Wavelet Features for Image Representation and Recognition ' In this paper, a novel two dimensional orthogonal wavelet features (2DOWF) method is presented for image representation and face recognition. The 2DOWF method derives 2D orthogonal wavelet (Gabor or Log Gabor) features in the feature extraction stage and then develops the cosine matrix measure for classification in the pattern recognition stage. 2DOWF method operates on the spatial structure of the pixels that defines the image. The wavelet transformed face images exhibit strong characteristics of spatial locality, scale, and orientation selectivity. These images can, thus, produce salient local features that are most suitable for face recognition. The two dimensional reduction PCA was used to detect noise, redundant features and form a representation in which these features are reduced. Analysis on the ORL dataset shows that the 2D orthogonal Log Gabor features are more suitable for face recognition than the 2D orthogonal Gabor features and the 2DPCA representation with an accuracy of 98.0% compared to 92.5% and 90.5%.
2d Orthogonal Wavelet Features
Log Gabor Features
Cosine Matrix Measure
Wavelet Transformed Face Images
Principal Component Analysis
Two Dimensional Reduction Pca