虹膜图像质量评价方法研究

Study on the Assessment Method of Iris Image Quality

虹膜识别中,图像质量除受到采集设备、环境光照、目标位置、目标状态等因素影响外,错误的图像分割也会导致图像质量下降,因此单一量度或较少量度都不能准确评价用于匹配计算的虹膜图像质量.基于此设计了可用虹膜区、虹膜大小、清晰度、扩张度等八个评价因子;研究了各评价因子所表征的图像质量特性;提出了利用图像分割参数计算各评价因子的方法;根据实验得到的各评价因子阈值以及图像分割可能的错误,对质量评价因子进行校准;最后融合校准后的因子形成虹膜图像质量评价值.实验结果显示所提出的方法可有效评价虹膜图像质量.

The quality of iris images is affected by characters of acquisition device, illumination condition, object position, and the status of the object. Incorrect segmentation may also lead to the low quality of iris area. Hence, the assessment method using one or two factors cannot cvaluate the iris image quality accurately. Based on above, our method exploits eight factors （ iris useable area, size of iris, sharpness, dilation, etc. ） to evaluate the iris image quality. The image characters represented by every factor are explained. The calculating methods of factors are proposed by use of picture segmentation parameter technology. In addition, the qudity evaluation factors are calibrated according to the thresholds getting from experiments and possibilities of the segmentation errors. The finale quality score is obtained by fusing these calibrated factors. The experiment results show that our method is efficient to evaluate the image quality.