Pixel-Level image Fusion Based on Fuzzy Theory

A new region based image fusion scheme is proposed. It is based on multiscale analysis. The low frequency band of the image multiscale representation is segmented into three kinds of regions by K-mean algorithm, which is used to determine the fusion rule and to achieve the multiscale representation of fusion result. The final image fusion result can be obtained by performing the inverse multiscale transform. The experiment demonstrates that the proposed image fusion method can illustrate better performance than exiting image fusion method.