摘要
复杂场景下红外弱点目标检测是红外目标识别的关键技术性难题之一。首先从概率论出发分析了背景因子存在的必要性。接着详细地论述了采用模糊集合中的隶属度理论和模式识别中的多分类器融合理论设计背景因子的过程。由于背景因子的引入对强干扰的复杂背景进行自适应地削弱,使得在各种背景下探测概率趋于均匀分布,最终实现了恒虚警率。
Under the complicated scene, the weak infrared target detection is one of the key technical problems in the infrared target recognition. This paper uses the probability theory to analysis the background factor＇s role. Then the membership theory on the fuzzy sets and the fusion theory of the multiple classifiers are introduced to design the back- ground factor. And the complicated background is adaptive weakened because of the background factor. At last the detection probability under various background tends to the uniformly distribution, and the constant false alarm rate is achieved.
出处
《激光与红外》
CAS
CSCD
北大核心
2009年第6期647-650,共4页
Laser & Infrared
关键词
目标探测
概率论
隶属度
多分类器融合
背景因子
target detection
probability theory
membership
fusion theory of the multiple classifiers
background factor
