Infrared face recognition based on blood perfusion and sub-block dct in wavelet domain

A novel method for infrared face recognition based on blood perfusion and sub-block DCT in wavelet domain is proposed in this paper. Firstly, thermal images are converted into blood perfusion domain by blood perfusion model to obtain consistent facial images to reduce the effect of ambient variations. Secondly blood perfusion data are decomposed using discrete wavelet transform. Then, the component of low frequency sub-bands is partitioned into sub-blocks, to which the DCT is further applied. Each face is represented by an effective feature combination extracted from the sub-blocks in DCT domain. Finally, the 3-NN classifier in Euclidean distance is utilized for recognition. The experiments conducted illustrate that the method proposed in this paper has better performance compared with traditional PCA, DCT and DWT + PCA methods.