Combined Contrast Enhancement Algorithm for High Dynamic Range Images

Image contrast enhancement is the process of improving the visual quality of an image by adjusting its brightness, color, and sharpness. When working with large bit raw images obtained directly from the matrix of the equipment, there are specific problems associated with a large dynamic range. The paper proposes a combined contrast enhancement method that can significantly improve the contrast of such raw images with a large dynamic range. In the combined method, highlight regions are softly clipped on the histogram using a clustering algorithm based on feature space partitioning and gamma correction of the clipped region. The clustering algorithm used does a good job of detecting the cutoff point, both in the presence of highlight regions and in the absence. The method also produces light border underlining based on Sobel filters. The well-known Contrast Limited Adaptive Histogram Equalization method is used to improve the histogram. In this case, a combination of transformations with different grid sizes is used, which allows to achieve much better results than when selecting one optimal transformation. These algorithms are described in detail and illustrated for comparison.