景象匹配算法适应性分析与仿真  

Analysis and Simulation of Adaptability on Scene Matching Algorithms

作　　者：汪洪桥[1] 赵宗涛[1] 蔡艳宁[1] 唐小丰[1] 

机构地区：[1]第二炮兵工程学院,西安710025

出　　处：《计算机工程》2005年第14期167-169,共3页Computer Engineering

基　　金：第二炮兵科研基金资助项目

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

关 键 词：景象匹配 基准图 实时图 匹配子图 加噪级数 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]