一种基于数据融合的身份识别方法

A method of Personal Identification Based on Data Fusion

针对单一生物特征身份识别由于传感器噪音及特征破损等缺陷导致识别率低的问题,从信息融合角度出发,提出一种基于语音和人脸的多生物特征身份识别方法。分别提取语音特征和人脸特征作为识别的依据,并用神经网络在特征层上进行融合识别。实验证明,该方法相对单一生物特征身份识别,在同等条件下具有更高的识别率。

Due to sensor noise, missing of characters, conventional recognition based on single-character gives a poor result, the pa- per presents an intelligent recognition system of muhi-character fusion, which is acoustic and face based on artificial neural net- work~ The experiment is shown to be superior to that of acoustic and visual subsystems.