图像识别技术在光电自动定向系统中的应用

Application of Image Identification Technique in Photoelectric Automatic Orientation Systems

将模式识别技术引入到光电定向之中,采取中值滤波、基于灰度分布特征的选择性灰度拓展、动态阈值计算和带修正的模板匹配等算法,有效地消除了不定光照和随机背景的影响,解决了航天器光电定向中光学图像识别定位的难题;该方法具有算法稳定、识别率高、定位准确的优点,为实现光电定向的自动化提供了有益思路。

Pattern Recognition technology in photoelectric orientation is introduced. It adopts medium filtering, grey scale expanding which based on the selectivity of grey scale distributing characteristic, dynamic threshold account, template matching and other arithmetic theories to eliminate the effect of uncertain illumination and stochastic background effectively, and solves the orientation problem of optical image manipulation in spacecraft photoelectric orientation triumphantly. Having characteristics of steady arithmetic, high identification and exact orientation, this method offers a profitable reference for the automatization of photoelectric orientation.