<![CDATA[In the digital era vulnerable to data security threats, Least Significant Bit (LSB) steganography emerges as an innovative technique for concealing sensitive text messages within digital images without altering visual perception. This study implements an adaptive LSB algorithm on 512x512 pixel grayscale and RGB images, employing random pixel selection based on a secret key to enhance resistance against statistical steganalysis. The embedding process converts text to binary and substitutes the LSB of the green channel, followed by reversible extraction achieving 100% accuracy under ideal conditions. Evaluation on standard datasets yields an average PSNR of 51.8 dB, MSE of 12.4, and embedding capacity up to 0.48 bpp, outperforming conventional LSB by 4-7% in imperceptibility and JPEG compression robustness (92% recovery rate). These findings affirm LSB's potential for practical information security applications, recommending hybrid cryptography integration to bolster resilience against adaptive attacks.]]>
Copyrights © 2026
