Hybrid histogram correction with adaptive gamma and multi-scale retinex for enhanced image processing

In the realm of Applied Artificial Intelligence, conventional image correction algorithms often face challenges when dealing with darker tonalities in images, leading to the loss of details, increased noise, and a lack of contrast and visual impact. To address these issues, this study proposes a novel hybrid histogram correction algorithm rooted in adaptive Gamma correction principles. This algorithm combines elements of the histogram equalization algorithm with the weighted distribution adaptive Gamma correction method, introducing a luminance weight to enhance precision. Building upon this foundation, the algorithm identifies pixels below a predetermined luminance threshold and enhances them using luminance augmentation techniques based on the Multi-scale Retinex (MSR) algorithm. Furthermore, to mitigate any excessive enhancement introduced by the MSR algorithm, the algorithm incorporates a corrective Gamma correction step, leading to the development of a comprehensive landscape color adaptation model. Experimental results using datasets relevant to urban landscapes demonstrate the effectiveness of the proposed algorithm in enhancing landscape imagery while preserving its natural essence. Moreover, the algorithm minimizes alterations to image chromaticity through advanced image processing techniques, ensuring the preservation of color fidelity.