Multiple iris image encryption using phase retrieval mask, QZ algorithm and umbrella map in hybrid transform domain

Iris image encryption is a significant research area in the field of image security. With the increasing use of iris biometric recognition technology in various applications, it is crucial to ensure the confidentiality, authenticity, and integrity of iris template images. In this paper, a novel multiple image encryption scheme using the QZ synthesis algorithm, chaotic umbrella maps, and unequal modulus decomposition algorithm in the hybrid transform domain is proposed. The proposed mechanism is non-linear and asymmetric in nature. The phase masks for encryption are generated using non-linear modified Gerchberg-Saxton phase retrieval algorithm, making the system resistant to chosen plaintext attacks, known plaintext attacks, and special attack. The use of the QZ synthesis algorithm provides a single channel encrypted image as output from multiple iris input images. Umbrella chaotic map algorithm is utilized for scrambling the pixels of the image. Experimental results demonstrate that the proposed encryption scheme provides a high level of security for encrypting multiple iris images simultaneously. MATLAB based simulations are conducted to validate and verify the performance of the proposed mechanism in terms of statistical analysis, key sensitivity analysis and attack analysis. The scheme has good efficacy and effectiveness against noise attacks, classical and special attacks along with higher sensitivity to various keys and parameters.