摘要
提高高分辨率SAR图像在复杂战场环境中的目标识别能力,对防御未来战争中来自地面目标的威胁具有重要意义。针对地面特定目标的大小、方位、旋转等变化以及强杂波背景给目标识别带来的严重影响,提出将目标的三维模型投影到二维平面,采用余弦傅里叶矩和瑞利分布的CFAR检测方法分别对其矩特征和峰值特征进行提取,利用级联组合分类器对目标识别进行建模分析,并通过试验验证该方法的有效性。结果表明:该方法实现了在特征维数高和姿态变化下的目标识别,而且无需额外增加对制导控制系统的开销。
It is of great significance for defending threats from ground targets in future wars to improve the abili ty of target recognition of high-resolution synthetic aperture radar（SAR） images in complex battlefield environ- ments. Aiming at the serious influence of the change of size, azimuth, rotation, and the strong clutter back ground on the specific ground target recognition, a method is presented. Firstly, the three-dimensional（3D） model is projected to a two dimensional plane. Then the moment and peak features are extracted by using the cosine-Fourier moment and constant-false-alarm-rate（CFAR） detector based on Rayleigh distribution, respectively. Finally, a cascaded combination classifier is used to model and analyze the target recognition. The effectiveness of the proposed method is verified by experiments. Results show that the proposed method achieves target recognition under the condition of high dimension features and attitude change, and it is not needed to increase the extra overhead of the guidance and control system.
出处
《航空工程进展》
CSCD
2017年第2期125-129,共5页
Advances in Aeronautical Science and Engineering
