系列虹膜图像质量评价

Study on series iris image quality evaluation

虹膜识别系统中,虹膜图像的质量是影响识别效果的重要因素。为了保证虹膜图像的真实有效性,首先要对输入的系列虹膜图像进行活体图像检测。然后进行质量评价,再根据评价结果选取质量好的图像。本文将虹膜图像质量评价分为4个部分:图像清晰度的评价、虹膜区域检测、位于虹膜区域的眼皮及睫毛检测,以便确定干扰因子的大小。最后进行综合质量评价,得出评价因子。试验结果表明,该方法能满足实时图像质量评价的要求。

The iris image quality is an important factor influencing recognition effectiveness in iris recognition systems. For ensure trueness and validity,the input series iris images should be firstly evaluated to judge whether it comes from live body. Then iris images with high quality can be picked out effectively based on quality evaluation. In this paper iris images quality evaluation is divided into four parts:definition evaluation,iris area detection,Eyelid and Eyelash detection in iris area to calculate interface factor. The comprehensive quality evaluation is computed to acquire evaluation factor. The experiment results show that this method makes possible to meet the requirement of a real—time images evaluation.