一种基于人工免疫的图像分割算法

图像二维熵分割,一直因耗时长而限制了实际应用。本文借鉴生物免疫思想,提出二维熵图像分割的人工免疫算法。在克隆选择算法中引入疫苗的免疫接种,用于优化最优分割阈值对的搜索过程。在遥感高分辨率图像上的实验显示,该算法不仅能准确搜索到最优阈值对,而且计算时间只有传统算法的1.8%。该算法也验证了人工免疫思想用于图像分割的可行性和有效性。

An Image Segmentation Algorithm Based on Artificial Immune

2D entropy method for image segmentation usually needs plenty of time, which limits its application. In this paper, by using biologic immune system for reference, we propose a new 2D entropy image segmentation algorithm based on Artificial Immune. Memory cells are imported as bacterins in clone selection algorithm for inoculation, in order to optimize searching process. Experiments on aerial images show good segmentation quality, and time is cut down to as only 1.8% as that of the traditional method. It proves our algorithm's feasibility and validity on image segmentation.