PROJECT TITLE :

Structure-Revealing Low-Light Image Enhancement Via Robust Retinex Model - 2018

ABSTRACT:

Low-light image enhancement ways based on classic Retinex model try to govern the estimated illumination and to project it back to the corresponding reflectance. But, the model will not take into account the noise, which inevitably exists in pictures captured in low-lightweight conditions. During this Project, we tend to propose the strong Retinex model, that additionally considers a noise map compared with the traditional Retinex model, to enhance the performance of enhancing low-lightweight pictures in the midst of intensive noise. Based on the robust Retinex model, we have a tendency to gift an optimization perform that includes novel regularization terms for the illumination and reflectance. Specifically, we use l one norm to constrain the piece-wise smoothness of the illumination, adopt a fidelity term for gradients of the reflectance to reveal the structure details in low-lightweight pictures, and make the first attempt to estimate a noise map out of the sturdy Retinex model. To effectively solve the optimization drawback, we provide an augmented Lagrange multiplier primarily based alternating direction minimization algorithm without logarithmic transformation. Experimental results demonstrate the effectiveness of the proposed method in low-lightweight image enhancement. As well, the proposed technique will be generalized to handle a series of comparable problems, such as the image enhancement for underwater or remote sensing and in hazy or dusty conditions.