首页 | 本学科首页   官方微博 | 高级检索  
     

基于Logistic回归模型的景像匹配算法性能评估
引用本文:王洋,徐刚锋,左东广.基于Logistic回归模型的景像匹配算法性能评估[J].红外技术,2008,30(9).
作者姓名:王洋  徐刚锋  左东广
作者单位:1. 第二炮兵工程学院101教研室,陕西,西安,710025
2. 第二炮兵装备研究院,北京,100085
摘    要:景像匹配算法性能评估的目的是通过科学客观地评估算法性能,选择满足应用需要的匹配算法及进一步改进算法.将自动目标识别(ATR)算法性能评估中常用的Logistic回归模型应用于景像匹配算法性能的定量评估,进一步深入研究了匹配试验所需的样本量和试验结果的定量分析方法,并通过对两种匹配算法的定量评估试验验证了方法的有效性.

关 键 词:匹配算法  性能评估  Logistic回归模型  样本量分析

Performance Evaluation of Image Registration Algorithm Based on Logistic Regression Model
WANG Yang,XU Gang-Feng,ZUO Dong-Guang.Performance Evaluation of Image Registration Algorithm Based on Logistic Regression Model[J].Infrared Technology,2008,30(9).
Authors:WANG Yang  XU Gang-Feng  ZUO Dong-Guang
Affiliation:WANG Yang1,XU Gang-Feng2,ZUO Dong-Guang1 (1.101 Department,the Second Artillery Engineering College,Xi\'an Shannxi 710025,China,2.The Second Artillery Equipment Research Institute,Beijing 100085,China)
Abstract:The purpose of the performance evaluation of image registration algorithms is to choose a registration algorithms and improving the algorithms by evaluating their performance scientifically. Logistic regression model generally used in ATR algorithms performance evaluation can also be used in quantitative performance evaluation of registration algorithms. Moreover, the amount of samples needed in registration experiments and quantitativel analysis methods for experimental results are deeply researched. Then ...
Keywords:registration algorithms  performance evaluation  Logistic regression model  amount of samples analysis  
本文献已被 CNKI 维普 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