A Deep Decomposition Network for Image Processing: A Case Study for Visible and Infrared Image Fusion

Abstract

Image decomposition is a crucial subject in the field of image processing. It can extract salient features from the source image. We propose a new image decomposition method based on convolutional neural network. This method can be applied to many image processing tasks. In this paper, we apply the image decomposition network to the image fusion task. We input infrared image and visible light image and decompose them into three high-frequency feature images and a low-frequency feature image respectively. The two sets of feature images are fused using a specific fusion strategy to obtain fusion feature images. Finally, the feature images are reconstructed to obtain the fused image. Compared with the state-of-the-art fusion methods, this method has achieved better performance in both subjective and objective evaluation.

Misaligned RGB-Infrared Object Detection via Adaptive Dual-Discrepancy Calibration

“…Furthermore, in order to prove the fusion effect of the spatial discrepancy calibration module on misaligned datasets, a large number of comparison experiments were carried out on the M3FD dataset with severe misalignment. Several existing excellent works, such as DDFusion [38], DIDFuse [39], TarDAL [22], CFR [25], GAFF [7], CFT [19], MFPT [32], and ProbEn 3 [8], are used to compare with our method. To ensure fairness, the input size of RGB and infrared images is set to 640 × 640, so all comparative experiments in this section did not use our size adaption process.…”

“…A high-level approach is intended to fuse RGB and infrared images into a new picture. Fu et al [38] decomposed RGB and infrared images into multiple sets of high-frequency and lowfrequency features by training a neural network, then added the corresponding features of the two modalities to form a fusion image. Zhao et al [39] implements a fusion network for RGB and infrared images based on an auto-encoder (AE).…”