RADON AND WAVELET TRANSFORM BASED APPROACH

Wavelet based representation has many advantages. According to psycho-visual research, there are strong evidences that the human visual system processes images in a multi-scale way. Converging evidences in neurophysiology and psychology are in consistent with the notion that the human visual system analyses image at several spatial resolution scales (Zhang 2004). An appropriate wavelet transform can result in robust representation with respect to illumination and expression changes and is capable of capturing substantial facial features, keeping computational complexity low.

This section presents Radon and Discrete Wavelet transform based framework for face recognition denoted as RDWT. The technique computes Radon projections in different orientations and captures the global directional features of the face images. Further, the wavelet transform applied on Radon space provides the directional multiresolution features of the facial images. For classification, the nearest neighbor classifier has been used. This algorithm is invariant to facial expression and illumination variations.