摘要： 建立了匹配仿真试验的贝努里(Bernoulli)统计模型，根据数字特征法，估计了基于序列图像匹配算法的匹配概率计算公式，证明了该模型的无偏性，并进行了精度分析。并给出了成功匹配的概率分布模型，定义了序列图像匹配算法的匹配概率，通过实例分析了序列图像匹配算法的匹配概率、图像序列数、单幅匹配概率之间的定量关系，为实际图像匹配系统算法、序列数目的选择提供了重要的理论决策依据。

关键词: 序列图像匹配, 匹配概率, 统计模型, 二项分布, 景象匹配

Abstract: This paper establishes the Bernoulli statistical model of match simulation experiments, estimates match probability calculation formula of image matching algorithm based on estimation method. With unbiasedness proved, it analyzes precision, describles probability distribution model of successful match in serial image match, and gives the definition of SIM algorithm match probability. The quantitative relationship among SIM probability, image serial number and single match probability is studied by an application example, which provides a theoretical decision-making basis for the choice of SIM algorithm and serial image number in aircraft scene match guidance using SIM technique.

Key words: serial image match(SIM), match probability, statistical model, binomial distribution, scene match

中图分类号: 

• TP18