A new multiple-image encryption method based on a light-field imaging system and the gravity model is presented in this paper. An inverse process of light-field imaging was used for multiple-image encryption. The method first combined multiple images into one light-field image. It was then encrypted using the gravity model, which further improved the security. The structural parameters of the light-field imaging system and the different objective distances of each original image could be used as part of the secret keys. A refocusing reconstruction algorithm for the light-field image was used in the decryption process. Experiments and analyses have shown that the proposed scheme could easily be accessed and provides a large key space and high sensitivity in the key space. Furthermore, it is robust against occlusion attacks. Compared with the repeated use of gravity model alone, the proposed method offers enhanced statistical properties and efficiency benefits. The proposed encryption can be achieved by a computational way, which can eliminate the negative effects of hardware limitations and aberrations.

提出了一种基于光场成像系统和重力模型的多图像加密新方法。光场成像的逆过程用于多图像加密。该方法首先将多个图像组合成一个光场图像。然后使用重力模型对其进行加密，从而进一步提高了安全性。光场成像系统的结构参数和每个原始图像的不同物镜距离可以用作秘密密钥的一部分。在解密过程中使用了针对光场图像的重新聚焦重建算法。实验和分析表明，该方案易于访问，并提供了较大的密钥空间和较高的密钥空间敏感性。此外，它对遮挡攻击也很强大。与单独重复使用重力模型相比，该方法提供了增强的统计特性和效率优势。可以通过计算方式来实现所提出的加密，这可以消除硬件限制和像差的负面影响。