Equalization of Images in Low Contrast Environments Through Adaptive Local Equalization for Enhanced Local Contrast Lighting

High-quality images are essential for several computer-vision applications. Hence, low-contrast images often pose significant challenges because they can cause failures in automated detection and analysis processes. This study presents an innovative strategy to improve image quality by segmenting an image into interleaved cell groups and processing each group independently using equalization. For each group, we construct a mesh of images with local equalization, then integrate it using a set of fusion functions. The method enables equalization not only based on the intensity histogram but also on the spatial properties of the elements, ensuring overall contrast enhancement across the entire image. Likewise, because the algorithm is deterministic, it is possible to estimate its latency and accelerate it reliably through hardware implementation. The performance of the proposed method was validated qualitatively and quantitatively against other state-of-the-art algorithms, demonstrating outstanding results.