摘要： 介绍了各种景象匹配算法,分析了实际应用中影响匹配性能的因素,并实现了算法适应性分析仿真平台。对两种经典的匹配算法,即MAD算法与Nprod算法,从内因和外因出发,理论分析了算法的复杂性,对噪声因素影响下的适应性进行了仿真试验,对算法的时间复杂性进行了定量比较,精确测量得出了正确匹配概率与加噪级数和信噪比之间的关系,最后得出了结论。

关键词: 景象匹配, 基准图, 实时图, 匹配子图, 加噪级数

Abstract: This paper introduces various scene matching algorithms and analyzes the factors influencing the matching performance in practicalapplications. It realizes an analysis and simulation platform for adaptability of the algorithms. From the internal and outer factors, a theoretic analysisof the complexity is gained about two scene matching algorithms, the MAD algorithm and the Nprod algorithm. On the platform, the two algorithmsare simulated with the factor of noise. The complexities of time of the algorithms are quantificationally compared, the relationship between the exactmatching probability and the noise adding series or SNR are enduced through the precise measurement. Also the paper draws a conclusion.

中图分类号: 

• TP391.41