液晶与显示, 2018, 33 (3): 228, 网络出版: 2018-05-29
基于GRT-CLEAN的高速旋转目标快速三维成像方法
Fast 3D imaging algorithm for high-speed spinning targets based on GRT-CLEAN
高速旋转目标 广义Radan变换(GRT)-CLEAN 网格剖分 三维成像 high-speed spinning target generalized Radan transform(GRT)-CLEAN mesh generation 3D imaging
摘要
针对传统高速旋转目标三维成像算法存在成像效果差、计算复杂度大、鲁棒性差的缺点, 本文提出一种加速的GRT-CLEAN高速自旋目标三维成像方法。采用广义Radan变换(GRT)与CLEAN技术相结合的方法进行目标三维特征提取, 实现高分辨精确目标成像; 采用“先粗网格, 后精确网格”的策略, 分两步对散射点目标进行估计, 降低计算复杂度, 计算复杂度从Ο(N×P×Q×T)降为Ο(10-4×N×P×Q×T), 大幅提高成像速度。仿真实验与数值分析验证了本文所提方法的有效性。实验结果表明, 本文所提成像在低信噪比和目标存在遮挡的情况下, 依然能对目标进行有效成像; 与传统的GRT-CLEAN成像方法相比, 本文所提成像方法大幅降低计算复杂度。
Abstract
In order to overcome the shortcomings of traditional 3D imaging algorithms for high-speed spinning targets, such as high computational complexity, poor imaging result and robustness, a fast 3D imaging algorithm based on GRT-CLEAN is developed. First, the 3D features of the target are extracted with the combination of generalized Radan transform(GRT) and CLEAN technology which could obtain a high resolution image. Then, the scattering targets are estimated with two separate grid sequentially (from coarse grid to fine grid), which could reduce the computational complexity and speed up the imaging process. The effectiveness of the current 3D imaging algorithm is verified by the simulation experiment and numerical analysis. Results show that the present 3D imaging algorithm based on GRT-CLEAN can obtain an effective image under the circumstances of low signal-to-noise ratio and no target occultation, and greatly reduce the computational complexity compared with the traditional GRE-CLEAN imaging method.
王会, 巨欢, 方阳, 李荣旭, 王保平. 基于GRT-CLEAN的高速旋转目标快速三维成像方法[J]. 液晶与显示, 2018, 33(3): 228. WANG Hui, JU Huan, FANG Yang, LI Rong-xu, WANG Bao-ping. Fast 3D imaging algorithm for high-speed spinning targets based on GRT-CLEAN[J]. Chinese Journal of Liquid Crystals and Displays, 2018, 33(3): 228.