摘要
针对遥感图像识别中的图像分割问题,运用群智能优化算法,以遥感图像为研究对象,在深入分析群智能缎蓝园丁鸟算法以及传统多阈值Otsu方法之后,提出一种改进的多阈值图像分割方法。该方法首先在种群初始化以及位置更新方面对标准缎蓝园丁鸟算法进行改进,改进以后的算法不易早熟。其次,通过转换,将最优阈值问题变为目标函数求解问题,以利用群智能算法的优点来快速找到最优阈值,从而完成阈值分割。最后,实际遥感图像分割结果表明,分割结果各项指标都较对比算法更优,抗噪性更好,速度更快。
Aiming at the image segmentation problem in remote sensing image recognition,using swarm intelligence optimization algorithm,taking remote sensing images as the research object,an improved multi-threshold image segmentation method is proposed after in-depth analysis of swarm intelligence satin bower bird optimization(SBO)algorithm and traditional multi-threshold Otsu method.This method first improves the standard satin bower bird optimization algorithm in terms of population initialization and position update,and the improved algorithm is not easy to be premature.Next,through conversion,the optimal threshold problem is turned into an objective function solving problem,so as to use the advantages of swarm intelligence algorithm to quickly find the optimal threshold,thereby completing threshold segmentation.Finally,the actual remote sensing image segmentation results show that all indicators of the segmentation results are better than those of the comparison algorithm,with better noise resistance and faster speed.
作者
孙振营
张文
Sun Zhenying;Zhang Wen(School of Information Engineering,Zhengzhou University of Technology,Zhengzhou 450044,China)
出处
《国外电子测量技术》
北大核心
2022年第2期33-39,共7页
Foreign Electronic Measurement Technology
基金
河南省科技攻关计划项目(212102210398,212102210401)资助
关键词
遥感图像
群智能算法
缎蓝园丁鸟算法
多阈值Otsu方法
remote sensing image
swarm intelligence algorithm
satin bower bird optimization algorithm
multi-threshold Otsu method
