摘要
针对高空无人侦察机遥感图像目标不明显、图像背景复杂等特点,提出了一种基于支持向量机的目标分类方法。将支持向量机应用于高空无人侦察机的遥感图像目标识别,并且对特征值的提取进行深入研究。实验结果表明,特征值输入的多分类器在识别性能上具有很高的识别率,对提高高空无人机的侦察能力具有重要意义。
According to the image of remote sensing of the high altitude unmanned scout is not obvious goal,image background complex,considering that having put forward a kind of goal classification method based on support vector machine. Using the support vector machine applied in the high altitude unmanned scout distinguishing image goal of remote sensing, and draw to carry out thorough research for feature value. The experiment result explains that based on the classfication ware of feature value input that this passage puts forward on identification performance have very high identification rate, and has important meaning for raising the reconnaissance ability of the highaltitude unmanned scout.
出处
《火力与指挥控制》
CSCD
北大核心
2010年第3期82-85,99,共5页
Fire Control & Command Control
基金
解放军射击学会"十一五"基金资助项目(03-14)
关键词
复杂背景
支持向量机
特征值提取
多分类器
complex background ,support vector machine ,extraction of feature value ,multi classifier
作者简介
唐克（1962-），男，安徽东至人，副教授，硕士生导师，研究方向：武器系统分析与仿真。
