【摘要】 对基于体全息的光电混合虹膜识别系统进行了研究,采用基于联合最优小波包基的特征图像相关识别方法实现虹膜的光学识别。生成联合最优小波包基的特征图像以压缩存储的模板图像库,降低后处理数据量。改进最优小波包基优选标准,使所选最优基获得最大识别能力。采用统计特征识别的后处理方法使光电混合识别系统具有更好的适应性。系统仿真结果表明,所提出的系统既具有光学运算和体全息相关的高并行性,又具有计算机数值计算的高精度,获得了良好的识别效果。更多还原

【Abstract】 Using the joint best basis selection based eigen-images correlation recognition method,the optical iris recognition is implemented in a volume holographic opto-electronic hybrid recognition system.In the hybrid recognition system,reference image bank is compressed by eigen-images generated by joint best wavelet packet basis selection algorithm.Both the number of images stored in photorefractive crystal and the time for post-processing are greatly reduced.The criterion to evaluate bases is modified for maxmizing the recognition capacity.Statistic feature recognition method is utilized for post-processing to make the system more robust.The processing speed of system is fast because of the high parallelism of optical computing and volume holographic correlation.With high accuracy of digital processing,the simulation shows that good recognition performance is achieved.更多还原