摘要
为了减小环境温度对红外人脸图像的影响,提高红外人脸识别系统的鲁棒性,提出了一种加权的红外图像线性归一化方法。首先,通过改进的0-1标准化得到图像中各点随温度变化的权系数,并利用这些权系数对图像进行温度补偿;然后,为了得到更加鲁棒的生物特征,通过血流模型将红外人脸图像转化为血流图;最后,用改进后的归一化相关系数和正确识别率对温度归一化后的图像进行检验。实验结果表明,温度归一化后的红外识别系统具有更高的性能。
In order to eliminate the effect of ambient temperature and improve the robust of the infrared facial recognition system,a weighted linear normalization method of infrared image is proposed in this paper.The variation coefficients of each point in the face are first obtained through the improved 0-1 standardization. To further improve the performance,the infrared thermal images are converted into blood perfusion images to get more stable biological features.Finally,the improved normalized cross coefficient and recognition rate are chosen to test the proposed method.Experimental results show that the infrared recognition system has better performance after the temperature normalization.
出处
《武汉理工大学学报》
CAS
CSCD
北大核心
2010年第20期1-5,共5页
Journal of Wuhan University of Technology
基金
国家自然科学基金(60665001
60862002)
关键词
红外人脸图像
0-1标准化
归一化
归一化相关系数
infrared face image
0-1 standardization
normalization
normalized cross coefficient
