首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于相对微分计盒法的人造目标分形检测
引用本文:杨绍清,赵晓哲,林洪文.基于相对微分计盒法的人造目标分形检测[J].光学技术,2006,32(Z1).
作者姓名:杨绍清  赵晓哲  林洪文
摘    要:对光电目标进行自动检测与识别是导弹末端智能制导的一项关键技术。基于图像分形维计算的相对微分计盒法,提出了一套实用的自然背景中人造目标的检测方法,给出了检测窗口与尺度选择的标准与参考值。大量实验证明,分形检测方法比传统的最大对比度检测方法稳定,是一种非常有前途的方法。

关 键 词:光电目标  图像分维  分形检测

Fractal artificial target detection basedon relative differential box-counting method
YANG Shao-qing,ZHAO Xiao-zhe,LIN Hong-wen.Fractal artificial target detection basedon relative differential box-counting method[J].Optical Technique,2006,32(Z1).
Authors:YANG Shao-qing  ZHAO Xiao-zhe  LIN Hong-wen
Abstract:Electro-optical tar get automatic detection and recognition is a key technology of intelligent guidance for a missile.Based on the relative differential box-counting method of calculating image fractal dimension,a practical approach is presented to detect an artificial target from natural scenes.The method analyzes the criterion for selecting the sizes of detection window and scale and their reference values.Vast computer experiments prove that fractal-based detection method gets stabler results than traditional maximum contrast method.Fractal-based detection method is a promising method.
Keywords:electro-optical target  image fractal dimension  fractal-based detection
本文献已被 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号