基于三维仿真平台的二维跟踪算法性能检测

对于现阶段昂贵武器不能重复外场实验进而影响光电跟踪经纬仪跟踪算法性能检测的问题，采用了一个模拟真实导弹飞行的三维仿真虚拟平台，通过改变参数模拟真实导弹飞行姿态等改变。主要利用坐标系转换和旋转映射技术解决了在三维仿真平台上检测二维跟踪算法性能的问题。两组实验结果表明:旋转映射技术在检测重心算法跟相关算法时稳定跟踪时平均帧频达到85 帧/s，并且比直接映射技术更节省计算时间，因此，已成功将二维跟踪算法应用于三维仿真平台，并可将该平台作为一个基准检测不同二维跟踪算法的性能。

2D tracking algorithm performance testing based on three-dimensional visual platform

Expensive weapons cannot repeat the experiment and affect the performance of detection and tracking algorithms, for this problem we use a virtual platform to simulate real missile flight and simulate different flight status by changing the parameters. It uses conversion between different coordinate systems and rotation mapping technology to solve the problem that detecting two-dimensional tracking algorithm performance in the three-dimensional simulation target. The experimental results show that: the average frame rate of rotation mapping techniques on stable tracking of gravity algorithm or correlation algorithm is 85 frames/s. And the time it takes less than direct mapping technology. The experiments show that the goal that detecting two-dimensional tracking algorithm performance in the three-dimensional simulation target is successfully achieved.