Directional projection based image fusion quality metric

In the past few decades, image fusion and its performance evaluation have attracted considerable research attention. However, it is still hard to objectively evaluate the fusion performance due to the diversity of image sources and the motivations for fusion. In this paper, we extend the work in image quality evaluation [8] to a novel metric for objective evaluation of image fusion. The input images and the fused image are firstly converted into local sensitive intensity (LSI) by Radon transform. We then employ the sensitive intensity for measuring how much information have been transferred from each source into the fused result by the difference of LSI. All the LSI pairs are finally incorporated into the expression according to the Weber-Fechner law. Experimental results demonstrate that our proposed metric outperforms other metrics while it is consistent with the subjective evaluation. (C) 2014 Elsevier Inc. All rights reserved.