Abstract:
The performance of a data fusion system must be better than that of single sensor system when the information of multisensor is effectively used. According to information theory, this paper discusses the classifying performance for a pattern recognition system based on multisensor neural network data fusion and proves the correctness. The conclusion is proved theoretically.
Key words:
Neural network,
Data fusion, Pattern recognition,
Valid Classifying information
摘要： 有效地利用多传感器的信息构建数据融合系统后，其性能优于单传感器系统。该文针对一种多传感器神经网络数据融合模式识别系统，对其分类性能进行了研究，以信息论的观点，从理论上证明了上述结论的正确性。
关键词:
神经网络,
数据融合,
目标识别,
有效分类信息
CLC Number:
ZHOU Kaili; KANG Yaohong. Study on Performance for Pattern Recognition Systems
Based on Neural Network Data Fusion
[J]. Computer Engineering, 2006, 32(17): 103-104，.
周开利;康耀红. 神经网络数据融合模式识别系统性能研究[J]. 计算机工程, 2006, 32(17): 103-104，.