A Relational Algebra Method to be Used to Decide Redundant Knowledge in Classification and Recognition Systems based on Multisensor Data Fusion

在基于多源信息融合的对象分类识别系统中,信源/传感器数量的增加往往使数据处理量急剧增长,不仅占用大量计算资源,提高运行成本,而且降低了系统可贵的实时性.本文提出用关系代数方法判定由信源传感器提供的知识有无冗余,从而精简和优化信源/传感器资源的配置和组合.