A Research Of Image Colorization By Non Local Total Variation Method

Colorization is a process of adding color to gray images.In this paper,we firstly introduce the popular methods recent years and then propose a new method for image col-orization by nonlocal total variation method in CB and YIQ color space.Lagrange multiplier method is considered to handle the sphere constraint of chromaticity in CB color space.By introducing an extra variable and using the dual version of nonlocal total variation,we split the colorization problems into some subproblems with closed-form solutions and get two iterative algorithms.Experimental results and comparisons demonstrate that the advantage of the proposed method is that the color edges can be better preserved than some of the existing methods,especially the total variation methods.And we also improve the method based on transfer image colorization and propose a non local block matching algorithm.