Based On Wavelet Transform Face Image Fusion Research

Infrared image can distinguish pretended target. In cloudy or foggy day it also easy can form images. In the contrary, the images which form with visible light are easily disturbed by weather. But spectrum resolution ratio of it is very good. Two kinds of these images can shoot in the same field of view. Then they can be fused with wavelet transform, and the information also can be strengthened.In the source image, different spatial resolution and frequency sub-image can be decomposed by the wavelet transform. According to the distribution of the data of high frequency sub image and low frequency sub image, a certain rules of fusion, form the effective fusion operator to calculate the original multi-source image fusion map in the proportion of the fused image, which is not only containing visible image feature but also containing infrared image feature.In order to evaluating the integration of image, text made quantitative statistics from the average, information entropy, cross entropy, distorting the extent of the correlation coefficient and so on.