PROJECT TITLE :

Gcs De color: Gradient Correlation Similarity for Efficient Contrast Preserving De colorization - 2015

ABSTRACT:

This paper presents a novel gradient correlation similarity (Gcs) live-primarily based decolorization model for faithfully preserving the appearance of the initial color image. Contrary to the conventional information-fidelity term consisting of gradient error-norm-primarily based measures, the newly outlined Gcs measure calculates the summation of the gradient correlation between each channel of the colour image and also the reworked grayscale image. Two economical algorithms are developed to resolve the proposed model. On one hand, because of the highly nonlinear nature of Gcs measure, a solver consisting of the augmented Lagrangian and alternating direction methodology is adopted to deal with its approximated linear parametric model. The presented algorithm exhibits glorious iterative convergence and attains superior performance. On the other hand, a discrete looking solver is proposed by determining the solution with the minimum perform price from the linear parametric model-induced candidate images. The non-iterative solver has benefits in simplicity and speed with solely many simple arithmetic operations, leading to real-time computational speed. In addition, it is very strong with respect to the parameter and candidates. In depth experiments underneath a variety of check images and a comprehensive evaluation against existing state-of-the-art strategies consistently demonstrate the potential of the proposed model and algorithms.