This repository represents the official implementation of paper titled： Preserving Local Features in Enhanced Images via Prior-Guided Histogram Equalization. If you find this repo useful, please give it a star ⭐ and consider citing our paper in your research. Thank you.
This paper proposes a modified histogram equalization technique that incorporates prior guidance to overcome limitations in traditional image contrast enhancement methods. By leveraging the local smoothing prior of the illumination component, our approach not only enhances image contrast but also preserves crucial local features that are often lost during nonlinear transformations. Comprehensive experiments on both grayscale and RGB images demonstrate that our method achieves significant improvements in contrast enhancement while maintaining image quality and structural features. Additionally, the proposed prior correction module can enhance the performance of various histogram equalization variants, broadening its applicability across diverse image enhancement tasks. This research contributes to advancing the field of image enhancement by providing a practical and effective solution for preserving local image characteristics during contrast enhancement.
Download following datasets:
LOL Google Drive
LOL-v2 Google Drive