ADVANCED AUDIO WATERMARKING WITH MULTIPLE IMAGE EMBEDDING FOR ENHANCED SECURITY

Abstract

Digital audio watermarking has gained a lot of popularity in the research community. Watermarking is a technique that uses several approaches (such as spatial domain, transform domain, machine learning, and deep learning) to conceal information signals into digital form Through the use of an audio file known as a watermark, the owner's identity can be concealed. Thus, digital audio watermarking is the technique of adding all of the data to the audio file without affecting the audio's audibility. This study describes a novel audio steganography approach based on paying a cryptographic hashing of audio samples in order to improve perceptual invisibility and security in the embedding procedure through a secret key. The suggested method uses a secret key to hash the most significant bit (MSB) of the audio samples for picking only certain samples for modulo embedding. The selection technique applied in this scheme by hashing the samples makes the watermark imperceptible to the unauthorized and reduces the probability of statistical detection by a huge amount. The bits of the different picture’s data are placed carefully into the LSB of the selected audio samples. Because of the careful placement of the watermark in the least significant byte (LSB) of the audio samples, the extracted watermark was found intact. A noticeable improvement in the quality of audio after the extraction process was observed.