Extended RGB2Gray conversion model for efficient contrast preserving decolorization

The RGB2GRAY conversion model is the classical and most popularly used tool for image decolorization. Recent researches have validated that optimally selecting the three weighting parameters in this first-order linear model has great potential to improve its conversion ability. A question is naturally raised that extending the parameter range will count for further improvement? In this paper, we present a simple yet efficient strategy to extend the parameter range for achieving such goal. In the extended model, the parameter range is extended to be [-1, 1] and the sum of three parameters is still constrained to be 1. A discrete searching solver is proposed by determining the solution with the minimum function value from the linear parametric model induced candidate images. Among the solving procedure, the newly presented vector p-norm of gradient correlation similarity measure is utilized. Extensive experiments under a variety of test images and a comprehensive evaluation against the state-of-the-art methods consistently demonstrate the potential of the proposed model and algorithm.