zhihongz/awesome-low-light-image-enhancement

Curated collection of datasets, deep learning methods (Retinex-based, histogram equalization, neural networks), benchmark papers, and quantitative metrics spanning raw sensor data to processed images across surveillance, autonomous driving, and microscopy applications. Organizes resources by methodological approach—learning-based, Retinex decomposition, and histogram equalization—enabling researchers to compare techniques and evaluate against standardized datasets like SID, LOL, and SICE. Serves as a structured index for practitioners selecting appropriate enhancement algorithms and evaluation protocols for low-photon-count imaging challenges.

1,780 stars.