摘要
随着国家低空空域的开放,防范和处置低慢小目标的干扰和破坏已经成为重大安保活动的世界性难题。上海天文台研制了一套快速搜索空中低慢小目标的光电系统。针对该光电系统,提出一种基于全景图像的低慢小目标的检测算法,以降低低慢小目标虚警率。首先利用Top-hat算子对低空全景图像进行高通滤波;然后对提取出的可疑目标进一步筛查判断,通过统计局部区域的天空、云朵和目标自身的灰度特性,设置门限以减少低慢小目标的误检率。对检测算法进行了试验验证,算法有效降低了形态学单帧检测算法虚警率,在低慢小目标搜索发现中取得了较好的工程应用效果。
An algorithm based on panoramic image is proposed to reduce the false alarm rate of low slow small target detection in this paper.First,the low-altitude panoramic image is filtered by the top-hat operator.Then the extracted suspicious targets are further screened and judged.By means of counting the gray-scale characteristics of the local area of the sky,cloud and the target itself,set threshold so that the false alarm rate could be reduced.
出处
《工业控制计算机》
2018年第5期66-68,共3页
Industrial Control Computer
基金
上海市科技成果转化和产业化项目(15DZ1160110)资助
关键词
低慢小目标检测
图像处理
复杂背景
形态学变换
low slow small target detection
image processing
complicated image background
morphlology transformation
