Denoising of Fingerprint Images using Q-shift Complex Wavelet Transform

Abstract

In this paper, we present a novel image denoising method using Q-shift with Dual Tree Complex Wavelet Transform (QDTCWT) for denoising fingerprint images. The DTCWT is an over complete wavelet transform with limited redundancy and generates complex coefficients in parallel using a dual tree of wavelet filters. But, low pass delay produces a Hilbert pair relationship between two trees. This is well addressed by Q-shift filters for improving orthogonality and symmetry properties in level 2 and below. QDTCWT have features like linear phase, tight frame, compact spatial support, good frequency domain selectivity with low sidelobe levels, approximate shift invariance, and good directional selectivity in two or more dimensions. This provides the QDTCWT basis mainly useful for de-noising purposes with high degree of shift-invariance and better directionality compared to the other traditional methods. The proposed algorithm has been designed in the MATLAB TM environment and tested in the fingerprint images obtained from the FVC2004 database for denoising. The performance and efficiency of the algorithm are estimated by calculating various quality metrics and compared with the advanced methods already practiced in fingerprint image denoising. The results of this study revealed that the QDTCWT algorithm is capable of producing high quality finger print images with greater fidelity, high robustness and accuracy over the other traditional denoising methods.