A Color Multi-Exposure Image Fusion Approach Using Structural Patch Decomposition

Abstract

A novel color multi-exposure image fusion approach is proposed to solve the problem of the loss of visual details and vivid colors. The proposed method is based on an image patch that is decomposed into three different independent parts: contrast extraction, structure preservation, and intensity adjustment. For structure preservation and intensity adjustment, which are very important, this paper exploits three weight measurements, i.e., local weight, global weight, and saliency weight. Using those weights, the final fused image is better guided by not only the exposure level of a single image but also the relative exposure level between differently exposed images. Then, the three parts are fused, respectively. The desired patch is reconstructed and fed back to the fused image. Unlike the conventional patch algorithms, the proposed approach preserves more details of the input sequence images due to the three weight maps. Experimental results demonstrate that the proposed approach successfully produces pleasing fused images with better color appearance and more texture details than other existing exposure fusion techniques.