提出了一种基于非线性尺度空间的非理想虹膜边界定位方法。这一方法通过具有尺度演化特性的非线性扩散消除小尺度的几何结构,保留虹膜的主要边界信息。此外将此扩散与多分辨率分析框架结合,从而实现了虹膜边界的快速有效检测。实验结果表明,与经典方法相比,所提算法能够有效消除上述因素导致的不利影响,精确提取了非理想虹膜的内外边界。
A nonlinear scale-space based localization method of nonideal iris boundaries was proposed. The method eli-minates small-scaled geometric structures while preserving large-scaled iris boundaries by nonlinear diffusions with the scale-evolution property. Moreover it combines such diffusions with the frame of multiresolution analysis, and achieves fast and effective iris boundary detection. Experimental results indicate that compared with the classical methods, the proposed algorithm extract nonideal iris boundaries accurately and efficiently with overcoming the disadvantageous in-fluences of the nonideal factors.