Evaluating IRIS Recognition Performance with DTC Wavelet Transform

Evaluating IRIS Recognition Performance with DTC Wavelet Transform

Anupam chouksey 1, Anoop Chaturvedi  2 Vikas Sakalle 3

1.              Phd Scholar, Computer Science & Engineering, LNCTU Bhopal (M.P)
2.              Professor, Computer Science & Engineering, LNCTU Bhopal (M.P)
3.              Professor, Computer Science & Engineering, LNCTU Bhopal (M.P

Abstract The dual-tree complex wavelet transform (DTCWT), a subgroup of the discrete wavelet transform, generates complex coefficients using a dual tree of ripple filters to obtain the real and imaginary parts. It provides low redundancy (2 m:1 for m-dimensional signals) while ensuring quasi-shift invariance and directional selectivity, which classical wavelet transforms lack. DTCWT maintains the properties of efficient computing and balanced frequency responses, making it ideal for multidimensional signals such as images. Applications include denoising, sparse coding and registration. In this study, DTCWT is applied to iris recognition, which has higher accuracy than existing methods by combining the features of DTCWT with low-pass and high-pass filters, thereby achieving best detection based on FRR and FAR performance.

Index Terms—DTCWT ; Iris; Image Processing ; Recognition; MATLAB.