低信噪比实孔径雷达机场目标图像匹配算法  

Registration of real aperture radar images of low SNR condition to optical images

作　　者：桑农[1] 胡冰[2] 

机构地区：[1]华中科技大学图像识别与人工智能研究所,湖北武汉430074 [2]华中科技大学图像信息处理与智能控制教育部重点实验室,湖北武汉430074

出　　处：《华中科技大学学报（自然科学版）》2008年第1期38-40,共3页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金重点资助项目(60135020)

摘　　要：提出了一种适用于低信噪比条件下的实孔径雷达机场目标图像匹配定位算法.通过对实孔径雷达图像中机场目标的检测识别,实现匹配定位.算法包括机场跑道检测及跑道端点定位两个阶段,分别采用了线特征滤波技术及目标背景的对比度信息.首先通过线特征滤波预处理和投影变换检测机场跑道所在的直线,然后利用机场在实孔径雷达图中的结构特征,根据机场及周围背景的对比度来定位跑道端点.实验结果表明:在成像条件较差的情况下,该算法仍能保证较高的匹配定位准确度.A method of registration for airfield detection in real aperture radar scene was proposed for its application under the circumstance of low signal noise ratio（SNR）. The location of airfield was based on detection and recognition in real aperture radar scene. The method consisted of two stages,including detection of runway lines and location of runway ends. They respectively adopted linear feature enhancement and contrast information of target background. Firstly located the runway lines through linear feature enhancement and radon transform. Then the method maked use of the strcture character of airfield in the aperture radar scene, and detected the runway ends according to the contrast information. The experiment showed that this method ensures matching precision under ill condition of real aperture radar imaging.

关 键 词：实孔径雷达 信噪比 机场检测 对比度 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]