Comparative Study of Multiresolution Analysis and Distance Measures for Face Recognition
Manisha Satone, G. K. Kharate · International Journal of Signal Processing Systems · 2013
In this paper we have compared 7 distance measures for face recognition using multi resolution analysis. Different wavelet families like Haar, Daubechies, Symlet, Biothogonal are used to test the results. Recognition experiments are performed using the database containing images of 400 persons. The experiments showed that the best recognition results are achieved using Manhattam distance measure on A2 and A3 subbands of symlets 8 wavelet.