摘要
信标光图像实时处理是完成空间光通信跟踪、捕获的一种过程。文中提出了一种改进的Otsu自适应阈值分割方法,对采集序列图像进行实时处理,提取潜在的信标光斑,后续图像处理引入感兴趣区读入技术,在感兴趣内对信标光图像进行处理。最后,比较了不同定位算法的跟踪结果,改进型Otsu克服了传统阈值分割信息不完备的缺陷,具有更强的抗噪能力、分割效果,实时性和定位精度均获得了满意的结果。
Beacon image processing in the spatial optical communication system is an important method for the tracking, acquisition and pointing（ATP） course. An automatic threshold selection method based on fast Otsu algorithm is proposed. The image processing is divided into two parts： the frist part is to identify potential target in a single frame, the second part is to detect beacon image in a ROI by using the improved Otsu algorithm. Finally, some experimental results for practical tracking are presented to show the effect of the differential algorithms. Experimental results show the Otsu algorithm overcomes the disadvantage of incomplete information for traditional algorithm,so it is beneficial to improve the ATP ability of system.
出处
《半导体光电》
CAS
CSCD
北大核心
2009年第2期260-263,267,共5页
Semiconductor Optoelectronics
