Aloka SINHA and Kehar SINGH^*

Photonics Group, Department of Physics, Indian Institute of Technology Delhi, New Delhi-110016, India

(Received March 7, 2003; Accepted June 26, 2003)

We outline a method to design a composite wavelet matched filter using the breeder genetic algorithm. The composite filter consists of a linear combination of wavelet and matched filters. Breeder genetic algorithm has been used to find the optimal weighting coefficients for the wavelet and the matched filters that constitute the composite filter. For the breeder genetic algorithm, the geometric mean has been used as the crossover operator and the creep has been used as the mutation operator. The performance of the filter is tested in the presence of additive white Gaussian noise. The performance of the composite wavelet matched filter has also been compared with that of the phase only filter and the classical matched filter. The designed filter can be used in an optical correlator to separate two classes of input images for fingerprint-based user identification in security systems.

Key words: matched filter, wavelet, breeder genetic algorithm, fingerprint identification, composite wavelet matched filter, Gaussian noise

^*Corresponding author. E-mail address: kehars@physics.iitd.ernet.in