A novel low-light enhancement via fractional-order and low-rank regularized retinex model

Most of existing low-light image enhancement approaches either fail to consider fine parts of the image or fail to consider intensive noise. To overcome these drawbacks, this paper proposes a new model called the fractional-order and low-rank regularized retinex model. Our model injects low-rank and fractional-order prior into a retinex decomposition process to suppress noise in the reflectance map and preserve the fine parts of the image. Our method estimates piece-wise smoothed illumination and noise-suppressed reflectance in turn, avoiding the residual noise in illumination and reflection maps that is usually present in alternative decomposition methods. After getting the estimated reflectance and illumination, we adjust the illumination layer to generate the enhancement result. Experiments on some challenging low-light images are presented to reveal the effect of our model and show its superiority over several state-of-the-arts in terms of enhancement efficiency and quality.