Adaptive Local Gamma Correction Based on Mean Value Adjustment

In complicated high dynamic scene, current gamma correction methods sometimes suffer from a loss of local details in certain regions of different luminance. In this paper, an adaptive local gamma correction method based on mean value adjustment is proposed, which is aimed at improve local details in both dark and bright regions in image. As the preprocessing, a global enhancement method is applied to improve the global contrast and then the proposed local gamma correction is undertook. Based on the Shih-Chia Huang's method, the proposed local gamma correction method has made improvement in two aspects and applied it to the local regions. For one thing, the average gray is regarded as one of the critical factors to determine the gamma value in the corresponding local region, which could adaptively improve the contrast both in dark and bright areas while preserve the original luminance of the regions that already own good contrast. For the other, followed by color correction, a novel histogram modification method aimed at avoiding the over-enhancement in smooth regions is used, which manager to maintain the histogram's original shape and suppress the large components in it at the same time. Experiments prove that the proposed method could get much more visual-pleasing and natural results.