Abstract: A new target recognition method based on wave coefficients is proposed, which entails expressing the scattering fields as a sum of cylindrical waves and using the distinctive coefficients as the feature vector. The investigation shows that the scattering coefficients are good target features. The coefficients are dependent on the incident angle, however, if the appropriate frequency range is chosen, the sensitivity to azimuth angle will be greatly decreased, leading to a smaller template database. The orientation of the target can be estimated to a certain degree of precision. It can be determined that wave-coefficients in the template database should be fetched to compare with the feature vector of the unknown target, thus effectively reduce the time of recognition process. Numerical simulations of four 2-D targets are conducted and ideal recognition performances are acquired, verifying the effectiveness of the proposed method.

Key words: target features, target recognition, wave-coefficients, 2-dimension target

摘要： 提出一种基于波系数的目标识别新方法, 此方法把目标响应表达成一系列波模的叠加, 波系数具有目标特异性, 因此可用作目标识别的特征矢量。这些波系数虽然与照射方向有关, 但是选取恰当的频率范围, 则波系数对角度的敏感性会大大降低, 目标模版数据库存储量会大大减少。在识别过程中可以大致预估目标的姿态角, 然后选择部分模版数据库与未知目标系数相比较, 可有效缩短目标识别时间。最后对4个二维目标进行了数值模拟, 取得了理想的识别效果, 证明了此方法的有效性。

关键词: 目标特性, 目标识别, 波系数, 二维目标