摘要
提出将光学相关识别应用于着陆点的选择中,以解决飞行器精确定点着陆问题。用"嫦娥一号"探测卫星传回的第一幅月球表面照片建立着陆点模型,通过尺度缩放模拟着陆过程中目标的尺度变化。利用优化折中综合鉴别函数(OTSDF)合成匹配滤波器,对尺度变化范围为2.5∶1的目标区域进行实时识别与定位,并根据相关识别结果计算了定位精度。仿真实验结果表明,光学相关的识别率高,定位误差远小于传统导航模式的定位误差,可应用于飞行器精确定点着陆中。
The method that uses optical correlation recognition in landing site tracking is proposed to achieve the aeroeraft precise landing. A high-fidelity simulation landing site environment is developed with the moon image obtained by Chinese first lunar satellite,and the scale variation of target with aerocraft descending is simulated by image zooming. Three of optimal tradeoff synthetic discriminant functions（OTSDF） filters are trained to detect the taget over a 2.5 ： 1 change in scale, and the position accuracy is calculated according to the correlation output. Simulation results show that the optical correlation recognition has high discriminant ahility and low position error,which is attractive in preeise landing site choice.
出处
《光电子．激光》
EI
CAS
CSCD
北大核心
2008年第12期1653-1655,共3页
Journal of Optoelectronics·Laser
基金
国家自然科学基金资助项目(60677041)
关键词
光学相关识别
精确着陆
优化折中综合鉴别函数
定位精度
optical correlation recognition
precise landing
optimal tradeoff synthetic discriminant functions
the moon image
position accuracy
