DWT, DCT and PBFO based Approach for Biometric Image Security
Astha Singh, Ankit Kumar, Malay Kishore Dutta · 2020
Concealment of biometric features widely attempted in the domain of watermarking of digital data. This draft is the representation of such an example in which fusion of iris (host image) and palm print image (watermark image) has been undergone into a robust and reliable watermarking scheme leaving unobjectionable distortion. These two biometric are logically fused by using the combination DWT (Discrete wavelet transform), DCT (Discrete cosine transform) and PBFO (Parallel bacterial foraging optimization). These 3 techniques are processed for the embedding of palm print image into the iris scan biometrics. The palm-print watermark preserves the ownership of host iris. A combination of both will be used for complete authentication. The quality check of the watermarked image and the watermark (after extraction) has been done by using standard approaches such as PSNR, NCC, BER, and MSE. The results say that the property of the iris image is unchanged after embedding the palm print in it and the image quality of palm print after the extraction process is also contented even after applying attacks. The rate of imperceptibility of host image after embedding is not only qualifying the standards but also conquer among the previous algorithm. The approach of this paper is to realize the concept of dual authentication which enhances the security of user's media.