论文部分内容阅读
本文采用前向迭代算法用在随机相位光学防伪中获得加密图象的Ｋｉｎｏｆｏｒｍ。该算法迭代过程中不再与被加密的原始图象进行比较，而是用加密图象振幅分布的均匀性控制迭代。这就克服了后向算法固有的局限性，在不增大每步迭代计算量的前提下，能够同时迭代获得一对实际存在的最佳加密相位列阵和加密图象的Ｋｉｎｏｆｏｒｍ。不仅如此，整个算法的收敛速度比后向算法快一个数量级，而且无论迭代结果的解密图象质量还是抑制噪声能力也都明显优于后向算法。
In this paper, the forward iterative algorithm is used to obtain an encrypted image of Kinoform in random phase optical anti-counterfeiting. Instead of comparing with the original image to be encrypted in the iterative process, the iteration is controlled by the uniformity of the amplitude distribution of the encrypted image. This overcomes the inherent limitations of the backward algorithm, and can simultaneously obtain a pair of Kincrys with the best cryptographic phase array and the encrypted image without increasing the computational complexity of each iteration. Moreover, the convergence speed of the whole algorithm is one order of magnitude faster than that of the backward algorithm, and it is obviously better than the backward algorithm whether the iterative result decodes the image quality or suppresses the noise ability.