Multi Exposure Image Fusion

High dynamic range images are images that have a ratio between their maximum and minimum intensities greater than that of display or capture devices. One method to create high dynamic range images is to combine a set of differently exposed shoots into a single image that has to be compressed before visualization. Compressing high dynamic range images is thus based on a so called high dynamic range map (HDR). In this paper we show that a solution to the problem can be obtained without the need for the calculation of HDR maps. To achieve this we use an image fusion method in the gradient domain. We start by calculating a gradient field which encompasses the details found in the exposure image series. This gradient field is then integrated to result in the final within dynamic range image. The resulting algorithm achieves high image contrast due to a non-linear thresholding of gradients.