Selective Colour Image Encryption Based on MSB and Sensitive Bits

Abstract

Efficient encryption algorithms for coloured images are crucial for maintaining device performance when working with limited resources. Traditional algorithms can be too demanding and ineffective for safeguarding data on such devices. Resource-limited devices, such as the Internet of Things, actively collect images and send them through the public internet. With the increased use of such devices, more secure, robust encryption algorithms need to be developed. To overcome these issues, this thesis proposes an algorithm that encrypts coloured images. This algorithm only selects the important parts within the image for the encryption process, reducing the resources required to encrypt images. In addition to selective encryption, we also implemented the proposed algorithm for full image encryption. This allowed us to perform evaluation metrics over selective and full encryption, gaining insights into the differences between performing full and selective encryption. From the evaluation results, our proposed algorithm achieved near-optimal values (average PSNR 8.7, 0.07 for MSSIM and 7.8 for IE) for the performed evaluation metrics, indicating a robust cryptosystem. Our proposed algorithm can be enhanced to include grey-scale images and develop a pseudo-random number generator.