【作者】 王娟； 慈林林； 姚康泽；

【Author】 Wang Juan~1 Ci Linlin~2 Yao Kangze~2 1 Department of Computer Science and Engineering, School of Information and Science Technology, Beijing Institute of Technology, Beijing, 100081, China 2 Information and High Technology Institute, Beijing, 100085, China

【机构】 北京理工大学信息科学技术学院计算机系； 信息高技术研究所；

【摘要】 提出了一种利用分形维和缝隙特征检测SAR图像舰船目标的新方法,分形维用来描述物体表面的不规则性,缝隙能够度量物体表面的质量分布,二者构造的检测特征矢量较为稳定,具有幅度不变性,以及不受目标姿态和类别影响的特点。通过分析SAR图像的分形特征,给出获取这两种特征的方法,并将提取的结果矢量用于舰船目标检测实验,结果说明了此方法的可行性和稳健性。更多还原

【Abstract】 A new method has been presented to detect the ship targets in SAR images based upon fractal dimension and lacunarity. Fractal dimension describes the irregularity and lacunarity measures the distribution of mass. The vector composed of these two features, is steady, and will not be effected by the pose and class of targets. The methods to estimate the feature values are discussed. Experiments are designed to apply these feature vectors to detect the ship targets. The results suggest the method is feasible and robust.更多还原