摘要
基于飞机的对称性特点 ,提出了一种航空图片中智能判别飞机类型的方法。该方法通过检测飞机的位置 ,以飞机的重心为矩心 ,根据矩不变原理 ,将飞机的中轴线旋转和坐标轴 Y轴重合 ,由飞机的轴对称特点求出 3个基本特征量 :飞机的长宽比 ;X轴方向投影值 ;Y轴方向投影值。利用匹配和投影相结合的方法可在航空图片中正确地识别型号不同、大小各异、位置任意的各种地面飞机。该方法识别特征少、计算量小 ,识别速度不足 1 s,在本实验中识别正确率达 1 0 0 %。
A new method which recognizes the aircraft models in aerial film according to the symmetry feature of aircraft is addressed. This method is based on moment invariants theory and selects CG(Center of gravity)as original point. With three basic invariable features including the ratio of length and width, the projected value of X\|axis, the projected value of Y\|axis, all kinds of aircrafts can be recognized. This method which can identify the target with little computation and high speed is proved effective. In the experiment, the ratio of aircraft recognition reaches 100 percent.
出处
《吉林大学学报(信息科学版)》
CAS
2002年第3期15-19,共5页
Journal of Jilin University(Information Science Edition)
基金
空军重点科研项目 (kj0 2 42 6)
关键词
识别
灰度图像
投影图
轴对称
航空图片
Recognition
Gray image
Projected image
Axial symmetry
Aerial film