基于运动形态学的掠地飞行器视频检测方法

 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 for Contour-chasing Aircraft Video Detection Based on Motion Analysis and Mathematical Morphology

A new method is proposed for contour-chasing aircraft video detection used in a irborne warning and control system (AWACS). The method is based on mathematical morphology, video motion-based analysis and statistics. The morphological topha t filter is used to extract suspicious objects (SO) in single frame firstly, the n the motion parameters of SO in adjacent frames are calculated. In the extracti on 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 a nd the target-like inference, some target-like interferences are removed durin g the statistics analysis processing. Experimental results obtained with the rea l outfield video image sequence demonstrate the effectiveness and robustness of the method.