日间恒星实时探测的视频图像处理  

Video image processing of real-time star detection in daylight

作　　者：冯小勇[1] 赵忠华[2] 刘新明[1] 

机构地区：[1]中国卫星海上测控部,江苏江阴214431 [2]上海交通大学仪器科学与工程系,上海200240

出　　处：《中国光学》2011年第6期622-628,共7页Chinese Optics

基　　金：总装备部重点试验技术研究项目

摘　　要：为了解决船载光学测量设备日间测星数量偏少的问题,研究了暗小目标电视探测技术,以保证惯导系统全天时不间断航向校准。在视频图像处理中,采用中值滤波方法提高信噪比,同时采用基于模糊理论的最大模糊熵阈值分割方法自行构建隶属度函数,自适应求取阈值,实现了星体目标检测与识别。实验结果表明:在原光学设备光学系统不变的情况下,日间可探测的恒星由大于等于3.0星等、小于等于5颗,提高到了大于等于4.5星等,大于等于250颗,满足了惯导系统全天时不间断航向校准的要求。该成果可以推广应用于电视弱小目标探测领域。To keep an inertial navigation system to correct its course errors continuously,a detection method is presented to solve the problem that few stars can be detected in daylight by the shipboard optical measuring equipment.The detecting method for the star body in daytime with strong background is studied.The median filter technology is used to improve the Signal to Noise Ratio（SNR） of an image,and the fuzzy entropy threshold method is used to creat its membership function and select the threshold of a gray-level image to detect star objects in daytime.The experimental result shows that the optical measuring equipment can improve the detecting ability from 3 MV to 4.5 MV and the amount of stars from 5 to 250 or more by this method.It meets the requirement of the inertial navigation system for correcting its course errors continuously.The method can be popularized and can be applied to the fields of dark and small target detection by televisions.

关 键 词：光学测量 惯性导航 最大模糊熵阈值法 电视探测 图像处理 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置] TN941.1[自动化与计算机技术—控制科学与工程]