Method for High Dynamic Range Imaging Based on Deep Guided and Self-learning br

Multi-exposure image fusion aims to fuse a series of images with different exposures for the samescene, and it is the main-stream method for high dynamic range imaging. To obtain more realistic results, aMulti-Exposure image Fusion Network(MEF-Net) based on deep guided and self-learning is proposed. Thisnetwork is designed to fuse any number of images with different exposures in an end-to-end way, and generatethe best-fused results in an unsupervised way. In terms of the loss function, an intensity fidelity constraint termand the weighted Multi-Exposure image Fusion Structural SIMilarity(MEF-SSIM) are introduced to improvethe fusion quality. Moreover, a self-learning method is adopted to fine-tune and optimize the pre-learned model,considering the fusion problem of two images under extreme exposure to mitigate the halo phenomenongenerated by fusion. Based on abundant testing data, experimental results show that the proposed algorithmout performs other mainstream methods in terms of both quantitative measurement and visual fused quality