A MULTI-DYNAMIC IMAGE STEGANOGRAPHY SYSTEM USING LEAST SIGNIFICANT BITS

Abstract

The study aims to design and implement a multi-dynamic image steganography system capable of embedding multiple data formats (colored and grayscale images, and text) into a single cover image, enhancing flexibility and security in data transmission. The system was developed using Python programming language and the Least Significant Bit (LSB) technique. Ten (10) colored and grayscale images, alongside text data, were used for system testing. The system’s imperceptibility was evaluated using the Peak Signal-to-Noise Ratio (PSNR) metric. A dual embedding and decoding process was established to handle both text and image files, enabling simultaneous concealment. The average PSNR value obtained was 39.21 dB, indicating good image quality and minimal distortion. Results also showed that the system achieved higher PSNR when embedding text compared to images due to smaller file sizes. The results demonstrate that the proposed multi-dynamic approach improves versatility and maintains perceptual image quality compared to conventional monodynamic methods. The developed system ensures secure, flexible, and imperceptible data concealment, supporting the advancement of intelligent steganography systems.