Region based Multi-focus Image Fusion using the spectral parameter Variance

Image fusion is the process of combining relevant information from two or more images into a single image. The resulting image will have more details than any one of the source images. In general, the Pixel based fusion methods are more sensitive to the noise present in input images. This leads to reduction in quality of fused Image. To face this challenge, a fusion method region based multi focus image fusion scheme is introduced based on image segmentation and the spectral parameter Variance in this paper. Firstly, an average image segment is calculated to perform the segmentation to input images. A shared region image is formed by using these partitioned images. The shared region is used to identify the input images. To perform the fusion process, a spectral parameter Variance is utilized. The proposed method produces better results on all types of images and proves to be the best method compared to all the traditional methods.