基于遗传算法的二维Otsu算法改进  

2D Otsu algorithm improvement based on genetic algorithm

作　　者：江禹生[1] 宋香丽[1] 任晶晶[1] 

机构地区：[1]重庆大学通信工程学院,重庆400044

出　　处：《计算机应用研究》2010年第3期1189-1191,共3页Application Research of Computers

摘　　要：为改善传统二维0tsu阈值分割算法处理图像时计算复杂度高、实时性差等缺点,将遗传算法应用到二维Otsu灰度图像阈值寻优中,并提出一种改进的自适应遗传算法。实验证明,新的算法对灰度图像有较好的分割效果,与传统算法相比,分割图像清晰,实时性也得到了明显的改进。The traditional two-dimensional Otsu threshold segmentation algorithm had the high computation complexity and poor real-time in processing image. To solve these problems, this paper adopted an improved adaptive genetic algorithm （IAGA） and used in the 2D Otsu gray image segmentation method. It was proved that the new method not only acquired a good effect, but also compared with the traditional 2D Otsu method, the segmented images were more clear and the real-time was advanced obviously.

关 键 词：二维OTSU法 阈值分割 自适应遗传算法 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]