摘要
针对多个特征指标的多传感器数据融合问题,将Fisher理论和多数投票法相结合进行数据融合来增加识别率。该方法首先通过Fisher理论得到多个判别函数,然后通过多数投票法继续对得到的判别进行分类得到最后的识别决策。该方法适合多个特征目标识别,计算简单,易于实现。
Aiming at the problem of multi-sensor data fusion with multiple characteristic indexes, we will combine the majority of voting theory and Fisher for data integration to increase the recognition rate. First, we get a number of Fisher discriminate function by Fisher, and then assort the result by the majority of voting theory to get the final classification of the identification of decision-making. This method is suitable for a number of features of target recognition, its calculation is vary simple, and it is easy to implement.
出处
《科技信息》
2009年第27期I0096-I0096,I0053,共2页
Science & Technology Information