Image Enhancement Based on Improved AGCWD Algorithm

In low-light and high-contrast environments, images often suffer from detail loss and uneven exposure, which seriously affects the visual detection performance. To this end, this paper proposes an image enhancement method that combines improved adaptive gamma correction (AGCWD) and multi-scale fusion. The method first dynamically adjusts the gamma value according to the global average brightness through the adaptive gamma correction module to enhance the dark areas while preventing the overexposure of highlights; subsequently, AGCWD is applied to the luminance channel in the YCrCb color space to balance the luminance through the weighted histogram mapping; and finally, the Laplace pyramid is utilized to fuse the enhancement results at the multi-scale level to achieve the dual optimization of details and overall luminance. Optimization. The experiments are carried out on a low-light dataset containing tables, vehicles, ships and complex backgrounds. Comparing with the traditional histogram equalization method, the results show that the proposed method improves the PSNR by 84.4% and reduces the MSE by 70.5%, which fully proves its significant advantages in image quality enhancement.