基于高斯曲面特征矩阵的扩展目标形状估计

发布时间：2019-01-29 06:06

【摘要】：针对传统方法难以准确估计扩展目标形状的问题,提出一种新的基于高斯曲面拟合的量测模型和基于高斯曲面特征矩阵的形状估计算法。首先,建立能反映目标真实形状的结构点,并产生多个高斯曲面,通过曲面叠加形成任意形状的量测空间分布模型;然后,根据高斯曲面拟合原理,用矩阵表示该拟合曲面主要区域的分布特征,并通过映射方程建立矩阵坐标与笛卡尔坐标的映射关系;最后,通过贝叶斯滤波体系更新拟合矩阵。与现有算法相比,本文算法不需要准确预设目标形状,在量测噪声较大的环境下,可以自适应的拟合目标真实形状。并且,在不需要预设目标形状方程的情况下,可以估计包括空心形状在内的任意不规则目标形状。实验结果表明,在目标初始形状参数不准确的情况下,本文算法正确估计了飞机形状、空心形状和集群目标形状,并且具有较好的扩展目标形状估计性能和较强的鲁棒性。
[Abstract]:In order to solve the problem that it is difficult for traditional methods to estimate the shape of extended target accurately, a new measurement model based on Gao Si surface fitting and a shape estimation algorithm based on Gao Si surface feature matrix are proposed. Firstly, the structure points which can reflect the real shape of the target are established, and several Gao Si surfaces are generated, and the measurement spatial distribution model with arbitrary shape is formed by the superposition of the surfaces. Then, according to the principle of Gao Si surface fitting, the distribution of the main area of the fitting surface is represented by matrix, and the mapping relationship between matrix coordinate and Cartesian coordinate is established by mapping equation. Finally, the fitting matrix is updated by Bayesian filtering system. Compared with the existing algorithms, this algorithm does not need to accurately preset the shape of the target, and can fit the real shape of the target adaptively under the environment of large measurement noise. Moreover, the shape of any irregular target, including hollow shape, can be estimated without the need for a preset target shape equation. The experimental results show that the proposed algorithm can correctly estimate the shape of the aircraft, the hollow shape and the shape of the cluster under the condition that the initial shape parameters of the target are not accurate, and it has better performance and robustness in the shape estimation of the extended target.
【作者单位】：江南大学物联网工程学院;
【基金】：国家自然科学基金(61305017,61304264) 江苏省自然科学基金(BK20130154) 江苏省研究生培养创新基金(KYLX_1125)资助项目
【分类号】：TN953

【相似文献】