基于分段自适应遗传算法的图像阈值分割  

Image thresholding segmentation based on piecewise adaptive genetic algorithm

作　　者：杜雯超 陈其松[2] 周莹[1] 

机构地区：[1]贵州大学大数据与信息工程学院,贵州贵阳550025 [2]贵州财经大学计算机信息学院,贵州贵阳550003

出　　处：《微型机与应用》2015年第3期58-59,62,共3页Microcomputer & Its Applications

摘　　要：自适应遗传算法是一种有效的寻优算法,本文首先对自适应遗传算法进行改进,提出分段自适应遗传算法,达到了防止早熟,加快寻优速度的目的。阈值分割是一种经典的图像分割算法,本文将利用改进的自适应遗传算法(分段自适应遗传算法)对图像分割。本文算法以最大类间方差比作为适应度函数,通过最佳阈值进行寻优,以信息熵和最大方差比作为评价标准对图像分割进行比较,实验证明基于分段自适应遗传算法的图像阈值分割算法能够达到较好的分割效果。Adaptive genetic algorithm is an effective optimization algorithm. Firstly the adaptive genetic algorithm was improved and the piecewise adaptive genetic algorithm is proposed, to prevent premature and accelerate the searching speed. Threshold segmentation is one of classical methods in image segmentation, in this paper the improved adaptive genetic algorithm（piecewise adaptive genetic algorithm） was used for image segmentation. Image segmentation based on piecewise adaptive genetic algorithm made the maximum between class variance ratio as fitness function, optimization of the best threshold. Finally, based on the information entropy and maximum variance ratio as evaluation criteria, comparing the results of image segmentation, image segmentation experiments show that the image segmentation based on adaptive genetic algorithm obtain good segmentation results.

关 键 词：图像分割 分段自适应遗传算法 最大方差比 信息熵 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置] TP391.41[自动化与计算机技术—控制科学与工程]