摘要
结合数学形态学、视频图像运动分析以及数学统计分析方法,提出了一种用于空中预警控制系统(AWACS)的小型掠地飞行器运动形态学视频检测方法。该方法首先应用形态学tophat滤波提取出视频图像中的可疑目标(SO),并求取SO的运动参数,利用真实目标和类目标干扰的不同运动特性,应用数学统计的方法对各运动参数进行分析以剔除部分类目标干扰,从而获得潜在目标(PT)的灰度分布和运动参数。在应用形态学tophat滤波器进行可疑目标提取中,一种新的金字塔型结构元的引入有效减少了虚警数量。外场试验表明,该检测方法能够有效抑制航拍掠地飞行器视频图像中的地物干扰,降低虚警目标的数量。
A new method is proposed for contour-chasing aircraft video detection used in airborne warning and control system (AWACS). The method is based on mathematical morphology, video motion-based analysis and statistics. The morphological tophat filter is used to extract suspicious objects (SO) in single frame firstly, then the motion parameters of SO in adjacent frames are calculated. In the extraction of SO, a new pyramidal gray structure element (SE) is used to help decreasing the false alarm. Benefited from the motion difference between the real target and the target-like inference, some target-like interferences are removed during the statistics analysis processing, Experimental results obtained with the real outfield video image sequence demonstrate the effectiveness and robustness of the method.
出处
《航空学报》
EI
CAS
CSCD
北大核心
2007年第1期151-155,共5页
Acta Aeronautica et Astronautica Sinica
基金
国家863高技术计划(2004AA823120)
国家自然科学基金(10376005)
关键词
微弱目标检测
形态学
tophat滤波器
运动分析
掠地飞行器
统计分析
空中预警控制系统
dim small target detection
morphology
tophat filter, motion-based analysis
contour-chasing air craft
statistical analysis
airborne warning and control system (AWACS)