Study of the influence of RGB and Lab color spaces on the performance of multifocus image fusion techniques

Abstract

In order to create a single image, multifocus image fusion brings together the important details and focused areas of the input multifocus images. Diverse camera depths of the field are used to capture these multi-focus images. Spatial domain focusing measurement has been used to introduce various multifocus image fusion algorithms. In this paper, we focus on the implementation of three multifocus color image fusion techniques: salience detection based multifocus image fusion (SDMF), salience detection using the technique of contourlet transformation (CT) and multi-scale guided filter (MGF) based fusion technique, using various color space models to improve the fusion result. The main objective is to examine whether the effects of R.G.B and L.A.B spaces have an influence on their recognition rate. We evaluated the performance of these techniques based on SSIM metrics, $\mathbf{Q}^{(\mathbf{AB} / \mathbf{F})}, \mathbf{L}^{(\mathbf{AB} / \mathbf{F})} \mathbf{N}^{(\mathbf{AB} / \mathbf{F})}$ and FMI. The experimental results on the test set show that the RGB space performs better than the LAB space.