因其高效的识别性能，虹膜特征码Iriscode被广泛应用于虹膜识别系统中。随着人们对生物特征安全问题的重视，如何从Iriscode重构虹膜图像以评判虹膜识别系统的安全性已成为一个研究热点。本文对Iriscode重构虹膜图像问题进行了分析，构建了虹膜重构优化模型，并提出两阶段的快速虹膜重构算法。将提取Iriscode的过程视为距离保持的降维过程，在第一阶段中根据待测Iriscode的近邻，采用局部线性嵌入算法重构出若干非精确的虹膜重构图像。第二阶段利用粒子群优化算法，将前一阶段获得的非精确虹膜图像作为初始粒子进行迭代搜索，以获得精确结果。实验表明，本算法获得的重构虹膜图像能够通过虹膜识别系统，达到伪造攻击的目的。相比于其他重构算法，在时间及伪造成功率上都有提高。

Because of its efficient identification performance, Iriscode has been widely used in iris recognition systems. With the emphasis on biosafety security, how to reconstruct the iris image from Iriscode to judge the safety of iris recognition system has become a research hotspot. In this paper, we analyze the problem of reconstructing iris images from Iriscode, construct an iris reconstruction optimization model and propose a two-stage fast iris reconstruction algorithm. We regard the process of extracting Iriscode as the dimensionality reduction process. In the first stage, some inexact iris reconstruction images are reconstructed according to the neighborhood of the given Iriscode using locally linear embedding algorithm. In the second stage, we use the particle swarm optimization algorithm and the inexact iris images obtained in the previous stages as the initial particle to perform the iterative search to search the accurate result. Experiments show that the reconstructed iris images obtained by the proposed algorithm can achieve the purpose of forging attack through the iris recognition system. Compared to other reconstruction algorithms, there is an increase in time consuming and accept rate of the reconstructed image.