一种基于四元数变换的彩色遥感影像检索方法  

A Remote Sensing Image Retrieval Method Based on Quaternion Transformation

作　　者：徐彦彦[1] 赵啸 李子君 XU Yanyan;ZHAO Xiao;LI Zijun(State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,Wuhan 430079,China;Fanruan Software Co.Ltd,Wuxi 450002,China)

机构地区：[1]武汉大学测绘遥感信息工程国家重点实验室,湖北武汉430079 [2]帆软软件有限公司,江苏无锡450002

出　　处：《武汉大学学报（信息科学版）》2019年第11期1633-1640,共8页Geomatics and Information Science of Wuhan University

基　　金：国家重点研发计划（2017YFB0504202）;国家自然科学基金（41571426）;武汉市应用基础研究计划（2017010201010114）~~

摘　　要：遥感影像常存在旋转和缩放等变化关系,这就要求检索时图像特征具有旋转和尺度不变等性质。目前,彩色遥感影像检索在提取这些特征时通常对色彩通道单独进行处理,导致3个通道之间的关系信息丢失,且没有充分利用数据本身所特有的几何特征,影响了检索精度。结合四元数与正交傅里叶-梅林矩的优点,提出了一种基于四元数变换的遥感影像检索方法。针对已有工作中存在的基于四元数傅里叶-梅林矩所提取的纹理特征只具有旋转不变性的问题,利用四元数正交傅里叶-梅林矩构造出具有旋转和尺度不变的纹理特征,并使用四其对图像进行边缘检测,得到边缘色彩图像并提取边缘色彩直方图,综合多种特征进行图像检索。实验结果表明,使用此方法进行遥感影像检索对于图像旋转和尺度变化具有良好的鲁棒性,检索性能明显提高。Remote sensing image retrieval requires that image features have the properties of rotation and scale invariance. Currently different color channels are often processed separately during the process of feature extraction, and unique geometric features of remote sensing image are utilized inadequately, which decrease image retrieval precision. A remote sensing image retrieval method based on quaternion transformation is proposed. Aiming at solving the problem of texture features extracted based on quaternion orthogonal Fourier-Mellin moments only have rotation invariance, we construct texture feature with rotation and scale invariance based on quaternion orthogonal Fourier-Mellin moments. We also use quaternion orthogonal Fourier-Mellin moment to detect image edges, get edge color image and extract edge color histogram,and integrate different features for image retrieval. The experimental results show that the proposed method has good robustness to image rotation and scale, and has better retrieval performance than other methods.

关 键 词：四元数 遥感影像检索 正交傅里叶-梅林矩 边缘色彩直方图 

分 类 号：P237[天文地球—摄影测量与遥感] P208[天文地球—测绘科学与技术]