Abstract: The binary image after image segmentation inevitablly contains noise and breakpoints, which brings a great difficulty to classification. Because binary image contains noise and breakpoints, Ant Colony System does not take full advantage of the characteristics of non-parametric curve with the random search location of initialization, large search scope and large amount of calculation of path selection. This paper puts forward Ant Colony System with Immune operator（ACSI）. It combines the advantages of Ant Colony System and Immune algorithm. Ant starts form the immune operator, while using local search strategies. Experiments  show that the algorithm has made the desired results both on effectiveness and efficiency.

Key words: immune operator, ant colony system, non-parametric curve

摘要： 图像分割后的二值图像中，难免会有噪声和断点，这对后期的分类等处理带来了很大困难。针对二值图像中含有噪声、断点，蚁群算法没有充分利用非参数曲线的特征，随机初始化搜索位置，搜索范围大、路径选择计算量大等，提出了免疫蚁群算法（ACSI）。它融合了免疫算法和蚁群算法的优势，从免疫算子开始搜索，同时采用局部搜索策略。实验证明，该算法在求解效果和效率两方面都取得了理想效果。

关键词: 免疫算子, 蚁群算法, 非参数曲线