Hierarchical Integrated ColorMatching in a Stereoscopic Image based on Image Decomposition

Abstract

Color discrepancies between the left and right image in a stereoscopic image cause many problems, including a reduction of the three-dimensional effect and increased visual fatigue. Thus, color matching in a stereoscopic image is very important for three-dimensional display systems. Therefore, a hierarchical integrated color matching method based on image decomposition is proposed for stereoscopic images. In the proposed method, global and local color discrepancies generated in a stereoscopic image are effectively reduced by histogram matching and illuminant estimation using image decomposition. The stereoscopic image is first decomposed into a base layer and several texture layers. Each decomposed layer is then matched using cumulative histogram matching and a multi-scale retinex algorithm. Lastly, inverse decomposition is applied to each layer to reconstruct the corrected stereoscopic image. Experimental results show that the proposed method has a better color matching performance in comparison with previous methods.