一种频域高分辨率遥感图像线状特征检测方法  

Linear Feature Detection for High-resolution Remotely Sensed Imagery in Frequency Domain

作　　者：周立国[1] 冯学智[2] 肖鹏峰[2] 马蔚纯[1] 

机构地区：[1]复旦大学环境科学与工程系,上海200433 [2]南京大学地理与海洋科学学院,江苏南京200091

出　　处：《测绘学报》2011年第3期312-317,共6页Acta Geodaetica et Cartographica Sinica

基　　金：国家863计划(2008AA12Z106);国家自然科学基金(40801166;41001234)

摘　　要：高分辨率遥感图像中线状信息过分细节化,相应的对特征的提取带来很大干扰。引入一种基于方向和频率特征的遥感图像频域线状特征检测方法,并通过傅氏变换将图像变换到频率域。在详细分析线状特征和谱线的关系、线状特征和图像频率之间关系的基础上,由分析得到的方向和频率的参数构造Gabor滤波器进行图像线状特征的提取。高分辨率遥感图像试验结果表明,该方法能较好地提取图像的线状特征。Information of linear features in high-resolution remotely sensed imagery is very prolific,consequently the immoderate specializations and influence of noise disturbances are serious,this made it difficult to detect and recognize the linear features of landscape targets.A method is discussed based on direction and frequency characters through designing frequency domain filter for the linear feature automatic recognition and detection,owed to Fourier transformation method transform the image into frequency domain.Analyzed the relations between the linear feature and its spectrum line,also the relations between the linear feature and its frequency character,according of the direction and frequency selected above the Gabor wave filter the image linearity feature extracting was constructed.The relevant extract experiments were tested with QuickBird image,the result indicates that the linearity characteristic extracted in this method is fairly good,it provided a reference for high-resolution remotely sensed imagery linear feature extracting.

关 键 词：线状特征 频域 GABOR滤波器 特征检测 高分辨率遥感图像 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]