复杂环境下异构传感器自适应景象匹配  

Research on the Heterogeneous Sensor Self-adaptive Matching Under the Complex Environment

作　　者：雷宝权[1] 周问天[1] 程咏梅[1] 李建博[1] 

机构地区：[1]西北工业大学,陕西西安710072

出　　处：《电力电子技术》2010年第8期62-63,87,共3页Power Electronics

摘　　要：提出了在复杂环境下异构传感器自适应景象匹配算法。应用异构传感器,将基准图常用的图像适配性指标引入到实时图评价体系中。采用多项指标对异构图像分别进行适配性评价,将实时图的每个评价指标作为一个证据,并给出各证据置信指派函数,通过证据融合完成异构传感器自适应的选择,进一步实施景象匹配。复杂环境下红外/可见光传感器自适应景象匹配的实验结果表明,该算法匹配概率高,可获得更好的匹配效果。Under the complex environment,the heterogeneous sensor self-adaptive matching algorithm is expatriated.By using the heterogeneous sensor,the common image matching of benchmark image is used in real-time imaging evaluation system.The multiple targets are applied for matching evaluation of heterogeneous image,and the belief assignment function of every evidence is given.The selection of heterogeneous sensor self-adaptive is accomplished by evidence amalgamation,and scene matching is implemented.The experimental results verify that the heterogeneous sensor self-adaptive matching algorithm has high probability and good matching effects.

关 键 词：异构传感器 自适应 证据融合 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]