Semantic pixel encoding visual secret sharing technique for balancing quality and security in color images

Abstract

<p>Color images are widely utilized across various domains, encompassing digital media and extending to critical applications in satellite and military arenas. As the significance of these images has grown, the need to protect their content from unauthorized access and potential threats has been underscored. Visual Secret Sharing (VSS) schemes have been proposed as effective mechanisms, with images being encrypted into multiple shares that, individually, offer no discernible information about the original content. Nevertheless, issues such as pixel expansion have been noted in traditional VSS methods, which result in increased complexity and a potential compromise in image quality. Maintaining impeccable image quality is emphasized, mainly since critical application decisions are often based on the clarity and accuracy of image details. The Semantic Pixel Encoding Visual Secret Sharing (SPEVSS) technique is proposed to address these identified challenges. A robust mechanism has been formulated through the integration of semantic pixel encoding with VSS, effectively countering pixel expansion while preserving the fidelity of the original image. As a result of this research, computational complexity has been significantly reduced, decryption methodologies have been made more efficient, and a more robust security framework for colour images has been established. The performance of the proposed SPEVSS shows the reconstructed images show the PSNR of 42 dB has been recorded in images processed, underscoring the method’s capability to balance security and optimal image quality.</p>