摘要
提出了基于支持向量回归机的红外热像仿真方法,该方法使用实测得到的场景红外辐射特性数据组成的训练样本集训练支持向量回归机,使得支持向量回归机具有预测能力,然后使用和物体表面划分面元数目相同的支持向量回归机计算仿真设定条件下的物体表面表观温度场数据,并利用三维可视化技术,将温度场数据和三维模型相结合,最终以运动车辆及建筑物场景为例,实现了对目标与背景构成的场景进行全方位红外热像实时仿真。
The infrared image simulation method based on support vector regression （SVR） was proposed.First,the SVR has the ability of forecast by using the trained samples which consist of testing data of the scene＇s infrared radiation characterization to train the SVR.Second,the surface temperature field of the object is calculated using some SVR the number of which is equaled with the number of segmentation elements of the object＇s surface.Third,infrared imaging real-time simulations of the mobile vehicle scene and the building scene are realized,combined the temperature field data with three-dimension models by using three-dimension visualization.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2010年第5期1323-1326,共4页
Journal of System Simulation
基金
安徽省红外与低温等离子体重点实验室开放基金资助项目(2007A010010X)
关键词
支持向量回归机
红外热像
运动车辆
建筑物
实时仿真
support vector regression
infrared image
mobile vehicle
building
real-time simulation
