An automated mean square image fusion for generating HDR image based on saturation

In this paper, a new method for generating high dynamic range (HDR) image is introduced. The proposed method can automatically fuses the input images of the same static scene with different exposure levels. Firstly, the saturation of the input image is solely used in this method. Since the saturation is similar to contrast, the weighing factor for blending the input images is determined by the mean square deviation. Then, the input images are blended into one output image as a linear combination. Therefore, the proposed method shows the efficient results without the halo effect and greater result when comparing to the existing method.