Performance evaluation of statistical measures for image fusion in spatial domain

Image fusion is the technique of combining useful information gathered from different visual sensors to present substantial image information. The fused image contains better information than any single source image. The statistical measures are used as quality gauge to identify the importance of the image or sub image. In general most of the images are affected by Gaussian noise. Statistical measures like uniformity and smoothness identify irregularities in the images in turn identify the sub images having less noise. Hence, uniformity and smoothness are explored in spatial domain for image fusion. All the superior quality sub images of the different source images are selected and combined to get the fused image. Quality assessment metrics are used to evaluate performance of the image fusion. The results indicate superiority of the smoothness measure in spatial domain than other measures for image fusion.