改进的强地杂波环境下车辆目标鉴别方法

Improved Identification Method of Vehicle Targets in Strong Ground Clutter Environment

摘要:

针对强地杂波环境下合成孔径雷达(synthetic apertu reradar, SAR)图像中的车辆目标鉴别问题,提出了一种基于变化检测量阈值分割和二维像素间隙度特征的车辆目标鉴别方法,利用像素散射强度变化等特征对SAR图像中的车辆目标进行目标鉴别可信度排序,并采用公开的车辆目标SAR图像数据集开展了仿真实验验证。仿真结果表明:利用二维像素间隙度特征向量计算得到的车辆目标鉴别可信度有3%~8%的提升。证明了所提方法的有效性。

Abstract:

Aiming at the identification problem of vehicle targets in synthetic apertureradar(SAR) image with strong ground clutter, an identification method of vehicle targetsbased on changeable detection threshold segmentation and two-dimensional pixel gap featureis proposed, The identification reliability of potential vehicle targets in SAR image is sortedby using the pixel scattering intensity variation, while simulation experiments are carried outbased on public vehicle targets SAR image datasets, which proved that the identificationreliability of vehicle targets counting by two-dimensional pixel gap features can be improvedby 3%-8%.The effectiveness of the proposed method is proved.