针对虹膜区域图像大小不同,难以实现虹膜的比较问题,提出了根据虹膜最小内边界周长确定角度分辨率和根据虹膜最小内、外边界半径差确定径向分辨率的方法,将不同分辨的虹膜图像转换为相同分辨率的矩形区域。结果表明,采用该方法得到的标准化维数对于虹膜的识别是有效的。
According to the fact that iris images cannot be compared for their different resolutions, a new method for normalizing iris region into fixed resolution by using minimum inner boundary resolution and minimum radius difference of inner boundary and outer boundary is presented. Results show that the normalization dimension achieved by the method is effective for iris recognition.