A coupled total variation model with curvature driven for image colorization
Zhengmeng Jin, Chen Zhou, Michael K. Ng · Inverse Problems and Imaging · 2016
In this paper, we study the problem of image colorization based onthe propagation from given color pixels to the other grey-levelpixels in grayscale images. We propose to use a coupled totalvariation model with curvature information of luminance channel tocontrol the colorization process. There are two distinct advantagesof the proposed model: (i) the involved optimization problem isconvex and it is not sensitive to initial guess of colorizationprocedure; (ii) the proposed model makes use of curvatureinformation to control the color diffusion process which is more effectivelythan that by using the gradient information. The existence of the minimizer of theproposed model can be shown, and the numerical solver based onconvex programming techniques can be developed to solve theresulting model very efficiently. Experimental results are reportedto demonstrate that the performance of the proposed model is better thanthose of the other color propagation models, especially when we deal withlarge regions of grayscale images for colorization.