Enhanced Image Steganography Using Customized Differences between the Neighboring Pixels

In the digital era, safeguarding sensitive information from unauthorized access is paramount. Image steganography, which inconspicuously embeds data into visual media, has emerged as a promising solution. However, balancing data embedding capacity with image quality remains a challenge. This study introduces a novel steganographic method that optimizes visual fidelity and payload capacity by utilizing customized differences between neighboring pixels within grouped blocks. Experimental results demonstrate the effectiveness of the proposed method, achieving a Peak Signal-to-Noise Ratio (PSNR) of up to 48.99 dB and a Mean Squared Error (MSE) as low as 1.20 on complex images like “Baboon.” Additionally, Structural Similarity Index Measure (SSIM) values consistently reach a perfect 1.00 across several test cases. The method also achieves a significant payload increase, embedding up to 119,496 bits in the “Resolution Chart” image. By leveraging the precise control over pixel differences, this study introduces a new paradigm in steganography, addressing the trade-off between capacity and quality while ensuring imperceptibility and high efficiency.