数据挖掘的热红外成像人脸识别研究

Research on thermal infrared imaging face recognition based on data mining

为了解决当前热红外成像人脸识别过程中存在的一些不足,提出了基于数据挖掘的热红外成像人脸识别方法。首先集热红外成像人脸,对原始热红外成像人脸进行预处理,消除一些不利因素对热红外成像人脸识别的不利影响,然后提取热红外成像人脸识别特征,最后采用数据挖掘技术设计热红外成像人脸识别模型,并与其他热红外成像人脸识别方法进行对比实验,结果表明,数据挖掘的热红外成像人脸识别准确率超过95%,单幅热红外成像人脸识别时间大约为10 ms,能够满足热红外成像人脸识别的实际应用要求。

In order to solve some shortcomings in the current thermal infrared imaging face recognition process, in this paper, a thermal infrared imaging face recognition method based on data mining is proposed. Firstly, collect the thermal infrared imaging face, preprocess the original thermal infrared imaging face, eliminate the adverse effects of some adverse factors on thermal infrared imaging face recognition, and then extract the thermal infrared imaging face recognition features, Finally, the thermal infrared imaging face recognition model is designed by using data mining technology, and compared with other thermal infrared imaging face recognition methods. The results show that the accuracy of thermal infrared imaging face recognition of data mining is more than 95%, and the face recognition time of single thermal infrared imaging is about 10 ms, which can meet the requirements.