Abstract: In view of the limitations of fast and precise iris location，an algorithm for image segmentation based on curve fitting is proposed.The main technique of the algorithm is that off-line acquire Fourier descriptor of boundary，as standard curve，of the clear iris-image，then search and fit the boundary of aim-image through standard curve for accurately and fast estimating the position of iris.The experimental result indicates，the algorithm can steadily realize that the goal is cut apart and drawn，and possess better robustness.

摘要： 针对多噪声情况下虹膜准确快速定位的困难，提出了一种基于局部曲线拟合的分割算法。该算法的基本思想是离线求出清晰虹膜边界曲线的傅立叶描绘子，然后利用标准曲线对目标图像中的边界曲线进行搜寻拟合，对虹膜位置进行判断，实现对其的快速准确定位。实验结果表明，该算法能够稳定地实现目标分割提取，具有较好的鲁棒性。