Iris structure has complex and plentiful textures which can be extracted as features for iris coding. We present a new representation of iris coding by using 2D wavelet. The proposed scheme of feature extraction is to use the multi-level coefficients of decomposition parts of image via wavelet. A feature vector consists of 5-level decomposition coefficients along vertical and horizontal direction. From these coefficients a compact binary code is generated.
|File Size||54.76 kB|
Windows Server 2008
|System Requirements||Matlab 7.0|