Hybrid Densely Connected Network for Multi-Exposure Image Fusion

Hao Zeng,Yulei Wang,Haoyang Yu,Meiping Song,Enyu Zhao,Tingting Tao
DOI: https://doi.org/10.1109/igarss52108.2023.10281494
2023-01-01
Abstract:Multi-exposure image fusion (MEF) technique is the most widely used method to obtain high dynamic range (HDR) images. Inspired by the recent successful application of Transformer in image processing, a hybrid dense connection network based on CNN and Transformer is proposed for MEF in this paper. Considering the importance of texture details to the multi-exposure image fusion task, shallow features containing rich texture details is also added to each dense layer, which are extracted by the pre-trained RepVGG. In addition, the dynamic weight calculation module is improved, so that different source images can obtain finer weight in the calculation of the loss function. Experiments are conducted on the dataset provided by MEFB, and both qualitative and quantitative comparisons show that the proposed method can achieve better results compared with the state-of-the-art algorithms.
What problem does this paper attempt to address?