基于混合蛙跳算法改进的OTSU遥感图像分割方法  

REMOTE-SENSING IMAGE SEGMENTATION METHOD BASED ON IMPROVED OTSU AND SHUFFLED FROG-LEAPING ALGORITHM

作　　者：路彬彬[1] 贾振红[1] 何迪[1] 杨杰[2] 庞韶宁 

机构地区：[1]新疆大学信息科学与工程学院,新疆乌鲁木齐830046 [2]上海交通大学图像处理与模式识别研究所,上海200240 [3]新西兰奥克兰理工大学知识工程与开发研究所

出　　处：《计算机应用与软件》2011年第9期77-79,105,共4页Computer Applications and Software

基　　金：科技部国际科技合作项目(2009DFA12870)

摘　　要：阈值的快速选取和噪声处理对图像分割起着至关重要的作用。针对遥感图像分割过程中阈值快速选取和噪声处理的问题,首次提出一种基于混合蛙跳算法优化改进的OTSU遥感图像快速分割算法。该算法首先对图像进行处理,引入一个邻域的空间和灰度相似测量因子来进行抗噪并且保护图像细节。再以最大类间方差作为混合蛙跳算法适应度函数,通过混合蛙跳算法的局部搜索和全局信息交换来快速确定图像分割的全局最佳阈值。实验结果表明,与传统OTSU图像分割算法及基本遗传算法改进的二维OTSU图像分割算法相比,该算法能更有效地去除噪声的干扰,算法运算效率更高。Fast selection of threshold and noise processing plays an important role in image segmentation.Aiming at these two issues in remote-sensing image segmentation process,an improved fast OTSU remote-sensing image segmentation algorithm based on the shuffled frog-leaping algorithm is proposed for the first time in the article.The algorithm first processes the image by introducing a similarity measuring factor incorporating both the neighbourhood spatial and gray-level relationships into the segmentation algorithm to perform anti-noise and image detail protection;then uses maximum between-cluster variance as the fitness function of shuffled frog-leaping algorithm,by applying shuffled frog-leaping algorithm in both local search and global information exchange,the fast determination of global optimal threshold of image segmentation is realised.Experimental results show that,compared with traditional OTSU image segmentation algorithm and the 2-D OTSU image segmentation algorithm improved from the basic genetic algorithm,the proposed approach is able to eliminate the noise interference more effectively with higher operation efficiency.

关 键 词：遥感图像分割 OTSU 邻域空间信息 邻域灰度信息 混合蛙跳算法 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]