摘要
以计算机视觉的不变性理论为基础,用飞机红外图像边缘中的角点反映飞机的几何结构.然后对飞机的角点应用变化,以构造出两种不变特征,实验表明这两种特征具有一定不变性.推导了Mahalanobis距离的平移、旋转和仿射不变性,并应用到空中目标的识别,实验表明这种方法效果良好。
Based on the invariant theory, corners on the edge of an airplane were used to describe the plane′s geometrical structure,and crossratio was applied for the corners to make two invariant features. The experiments show that the two features invariant in some scopes.It was deduced that the Mahalanobis distance is invariant in translation, rotation and affine deformation.So it can be used to recogniz air targets. The experiments show that the method has good results and stable performance.
出处
《红外与毫米波学报》
SCIE
EI
CAS
CSCD
北大核心
1997年第1期39-44,共6页
Journal of Infrared and Millimeter Waves
基金
国防预研基金