Multi Focus Image Fusion Using Statistical Approaches

Abstract

Pictures of the long focal lengths cameras, suffer from the problem of limiteddepth of field. Consequently, it is often difficult to obtain good focus for allobjects in the picture. In this work, the problem of multi focus is solved by takingseveral pictures with different focus points, and then fusing them together to forma single image using image fusion method. Contrast of the source images is usedas a measure to select the proper pixels (pixels with high focus) to obtain thefused (output) image. This work suggests using three statistical measures ofcontrast to calculate the pixels gradient magnitude. Those three measures areapplied and compared with the traditional method of gradient magnitude measures(Sobel and Perwit) in Multi focus Image Fusion. The results show that themodified methods gave better results than Sobel and Perwit methods.