Concealing Fingerprint-Biometric Data into Audio Signals for Identify Authentication

Numerous security issues are raised through the transmission and storage of biometric data due to its high sensitivity and extremely crucial purpose. However, it might be impossible to recover if lost, counterfeited or hacked, thereby ruining the general aim of securing it. In this paper, an 8-layered feature enhancement algorithm is proposed. A centroid based semi-fragile audio watermarking is used to conceal the enhanced fingerprint biometric data into audio signals. The algorithm starts by encrypting our enhanced/extracted fingerprint template and converting it into binary bits prior to watermarking. It hence proceeds with computing the centroid of each audio frame and once more encrypting the obtained watermark using a Logistic map operation. DWT and DCT are performed on the sub-band which carries the centroid of audio thereby embedding the encrypted watermark bits into the domain signal. ODG and SDG analysis are performed on both the original and watermarked signal and the results signify the efficiency of the proposed scheme. Moreover, some signal processing operations are finally carried out on the watermarked signal and the outcome was intriguing as all attacks are counteracted.