基于极化交叉熵和Yamaguchi分解的飞机目标检测方法

Airplane Detection Algorithm of Polarimetric SAR Image Based on Polarimetric Cross Entropy and Yamaguchi Decomposition

给出了一种基于极化交叉熵和Yamaguchi分解相结合的飞机目标检测方法。首先计算极化SAR图像中各个像素点与平面、左螺旋体、右螺旋体三种基本散射体的相似性参数,并利用相似性参数构造极化交叉熵;然后采用Yamaguchi分解方法提取偶次散射分量功率;最后结合极化交叉熵与偶次散射分量功率构造检测特征量,并进行阈值判别提取飞机目标。利用美国UAVSAR和美国AIRSAR系统采集的全极化实测数据对算法进行实验,结果表明,该算法能够有效的检测出飞机目标,并且虚警较少。

In this paper,a new airplane detection algorithm of polarimetric SAR image based on polarimetric cross entropy and Yamaguchi decomposition is proposed. First of all,the similarity parameters are computed between each pixel of PolSAR image and three canonical scattering mechanisms which contain surface scattering mechanism,left-helix scattering mechanism and right-helix scattering mechanism. Further,construct polarimetric cross entropy based on the similarity parameters. Then,Yamaguchi decomposition is adopted to extract double-bounce scattering component power. At last,detection feature combining polarimetric cross entropy with double-bounce scattering component power is proposed,and the technology of threshold segmentation is used to achieve airplane detection. Multi-look fully Pol SAR datasets acquired by U. S. UAVSAR systems and U. S. AIRSAR systems are used to test and verify the efficiency of the new algorithm. The experimental results show that the novel method can detect airplane effectively. Besides,the detection result’s false alarms are less.