Skip to Main Content
Image fusion is a proven value adding technique for image analysis. Automated image fusion aims to give the fusion system the ability to select, analyze and evaluate fusion-worthy images. This paper examines the evolution of present techniques used for assessing quality of image fusion operators. It also presents an algorithm that objectively evaluates the realism of saliency functions used in image fusion quality measures. Most image fusion quality metrics depend on estimating the amount of information transferred from each source image into the fused image. This algorithm rebuilds the fused image using the estimated information from each source image and compares it to the original fused image.