摘要
为了检测复杂海面背景中的舰船目标,提出了一种基于图像序列区域混沌特征的目标检测新算法,算法利用小数据量法计算图像序列区域的最大Lyapunov指数,分析运动目标存在时背景信号混沌特征的变化,并利用混沌特征的变化差别检测淹没在混沌背景信号中的目标信号,最后对100帧图像进行了目标检测。计算结果表明:新算法检测率为100%,虚警率为5%,检测结果优于利用统计分析方法的结果。
In order to detect ship target on complex sea surface background, a new algorithm based on the block chaos feature of image sequence was proposed, the block largest Lyapunov exponent of image sequence was calculated by using small data sets, the change of background chaos feature was analyzed when moving target existed, ship target signal submerged by background signal was detected based on the change difference of chaos feature, and 100 frame images were used to detect ship target. Experiment result shows that the detection rate of the proposed algorithm is 100%, its false alarm rate is 5%, so the algorithm is superior to statistical analysis method. 1 tab, 1 fig, 16 refs.
出处
《交通运输工程学报》
EI
CSCD
北大核心
2009年第1期73-76,共4页
Journal of Traffic and Transportation Engineering
基金
国家自然科学基金项目(60572160)
关键词
交通信息工程
图像序列
混沌特征
小数据量法
目标检测
traffic information engineering
image sequence
chaos feature
small data sets
target detection