【摘要】 传统的二维Otsu阈值分割算法采用穷举搜索法搜寻阈值向量。与此不同，本文提出了一种二维类间方差阈值分割的快速迭代算法，用迭代的思想解决原始二维Otsu方法计算复杂、实时性差的问题。文中导出了迭代算法的公式，给出了算法流程。实验结果表明，与二维Otsu原始算法及其他两种快速算法相比较，本文提出的二维Otsu快速迭代算法分割结果准确，实现简单，其运行时间仅为原始算法的0.4%左右，大大减少了计算量和存储空间，是一种快速有效且实时性好的图像阈值分割算法。

【关键词】 图像分割； 二维类间方差； Otsu阈值； 快速迭代考试大:/

A fast iterative algorithm for image segmentation based on 2D

maximum betweencluster varianceWU Yiquan, WU Wenyi, PAN Zhe

（College of Information Science and Technology, Nanjing University of Aeronautics and Astronautics,Jiangsu Nanjing 210016, China）

Abstract: The traditional twodimensional (2D) Otsu threshold algorithms for image segmentation always use exhaustive searching method for the best thresholds. In this paper, a fast iterative algorithm based on 2D maximum betweencluster variance is proposed in order to improve the performance and efficiency of the original 2D Otsu threshold algorithms. The iterative formula is deduced and the algorithm flow chart is given in the paper. Experimental results show that the proposed algorithm has a good segmentation result pared to the original 2D Otsu algorithm and the other two fast methods. It can well reduce the storage space and the running time which is only 0.4% of that of the original method. Therefore, it is a fast and effective image segmentation algorithm with a good realtime quality.

Key words: image segmentation; 2D maximum betweencluster variance; Otsu threshold; fast iterative

引言