基于血流图双树复数小波域傅里叶变换的红外人脸识别方法

提出了一种基于血流图与双树复数小波域傅里叶变换的红外人脸识别方法．首先利用血流模型把温谱图转换成血流图，然后将人脸血流图进行一级双树复数小波分解，保留分解后的4个低频子带并分别进行傅里叶变换，得到每个低频子带的特征矩阵，分别计算4个子带的欧氏距离并进行简单的加法融合，再用三阶近邻分类器得到最终的识别结果．为了减小算法的时间复杂度，我们对血流模型进行简化．实验结果表明，本文所提的方法有较好的识别结果．而简化的血流图相比原模型的识别率没有明显的下降，甚至某些情况下还稍高于血流模型，说明本文的方法是有效的．

Infrared Face Recognition Based on Blood Perfusion Model and FFT in Dual2Tree Complex Wavelet Domain

An efficient method for infrared face recognition by blood perfusion model of human face and FFT is proposed. Firstly, Each infrared face image is converted into blood perfusion data by blood perfusion model to obtain consistent facial images without effect of ambient variations. Secondly, blood perfusion data are decomposed using one scales＇ Dual2Tree 2-D Complex discrete wavelet transform. Then, four tow frequency subbands obtained after transforming are further transformed via Fourier Transform （FT）. The features extracted from the reserved subbands in FT domain are used for recognition computing four subbands＇ Euclidean distance and fusing them using simple add. then using the 3-nn classifier in the Euclidean distance to obtain the final recognition results. In order to reduce time complexity of our algorithm, the blood model is modified, by principles of infrared imaging and biological heat transfer and temperature information. The experiments conducted illustrate that the method proposed in this paper has better performance. While the recognition rate wasn＇ t decrease based on modified blood model compared to blood model obviously and have even lightly improved in some cases,it shows that our method is efficient.