Proceedings Article | 15 November 2022
KEYWORDS: Image encryption, Skin, Optical image encryption, Image quality, Image fusion, Optical coherence tomography, Biometrics, Computer security, 3D image processing, Image processing
Since the human body has its own unique biological characteristics, such as fingerprints, faces, voices, etc., the biological characteristics of the human body are widely used as a key for image or information encryption to improve the security of the system. However, the biological characteristics of the human body are abused, resulting in reduced safety. In addition, the double random phase template is widely used as the key in optical image encryption technology. However, the linear nature of the double random phase template is vulnerable to attack. Therefore, there is an urgent need to develop a new type of key with high security and not easy to be attacked. In this article, we propose a three-dimensional skin thickness key based on fingerprint guidance. The key generation algorithm includes three steps. First, the convolutional neural network is used to segment the upper and lower boundaries of the epidermal layer in optical coherence tomography (OCT) images of the fingertip skin cross-section, and the deep learning algorithm is used to extract the maximum intensity projection (MIP) image of the internal fingerprint at the fingertip skin epidermis-dermis junction (DEJ). Secondly, by locating some areas in the MIP, the thickness of the upper and lower borders of the skin of part of the fingertips is calculated and converted into a thickness map. Finally, based on the characteristics of the internal fingerprint, the thickness map is selected as the key to encrypt the optical image. The experimental results show that the thickness of the skin of each person's fingertips is different under normal conditions, and the thickness of the skin can be encrypted as important biometric information. Numerical simulation verifies the feasibility, security and robustness of the encryption scheme.
