基于快速模型匹配的飞行目标三维姿态角估计算法

Estimating Algorithm of 3D Attitude Angles of Flying Target Based on Fast Model Matching

采用动态模型匹配的基本原理,提出了一种基于二维图像序列的飞行目标三维姿态角估计算法。算法根据目标机动特点和姿态角约束关系,建立了降维动态模型库,在保证模型库完备性的同时减小了规模。根据相似性矩阵的分布规律提出了十字搜索法,加快了最佳匹配模型的搜索速度。仿真表明,该算法具有较好的准确性和实时性,对提高机动目标跟踪性能具有一定的理论意义。

Based on the principle of dynamic model matching,an estimating algorithm of 3D attitude angles of flying target was proposed based on a 2D image sequence.Lower dimensional dynamic model library was established according to the feature of maneuvering target and constraint relation of attitude angles,which insured the completeness of model library and simultaneously reduced the scale of it;Cross searching method was proposed according to the distribution of similarity matrix,which accelerated the searching of optimal matching model.The test shows that the proposed algorithm has better precision and real-time performance,and is theoretically significant to improve the performance of maneuvering target tracking.