6 November 2024 in IEEE Transactions on Geoscience and Remote Sensing
Remote sensing images can accurately display the electromagnetic attribute distribution of various ground objects and can be used in many applications. Nevertheless, stripe noise frequently degrades the quality of the captured images. Most current destriping studies are capable of removing the regular stripe noise. However, their outputs often produce a stripe residual or oversmoothing effect when a complex stripe noise exists. To overcome this limitation, in this paper, we propose a variational destriping model based on nonconvex regularization and variable weights. First, our model constrains the stripe sparsity using the normalized ϵ-penalty function, and converts it into weighted ℓ 1 norm by ℓ 1 majorization method. Second, unlike previous models that use scalar smoothing weights to control the output smoothness, we propose to design adaptive vector weights to constrain the smoothness. Moreover, the proposed model introduces the region weight to deal with the extreme regions. In view of the region decomposition technique, we transform our optimization model into a series of 1D weighted ℓ 1 problems in different directions with a linear time solver. Then, an improved alternating direction method is employed to provide fast convergence. The experimental results show that our method achieves a better performance on image detail preservation, complex stripe removal and a faster convergence than the state-of-the-art methods.
27 February 2026 in IEEE Transactions on Multimedia
Unmanned aerial vehicle (UAV) target tracking based on thermal infrared imaging has been one of the most important sensing technologies in anti-UAV applications. However, the infrared UAV targets often exhibit weak features and complex backgrounds, posing significant challenges to accurate tracking. To address these problems, we introduce SiamDFF, a novel dynamic feature fusion Siamese network that integrates feature enhancement and global contextual attention knowledge distillation for infrared UAV target (IRUT) tracking. The SiamDFF incorporates a selective target enhancement network (STEN), a dynamic spatial feature aggregation module (DSFAM), and a dynamic channel feature aggregation module (DCFAM). The STEN employs intensity-aware multi-head cross-attention to adaptively enhance important regions for both template and search branches. The DSFAM enhances multi-scale UAV target features by integrating local details with global features, utilizing spatial attention guidance within the search frame. The DCFAM effectively integrates the mixed template generated from STEN in the template branch and original template, avoiding excessive background interference with the template and thereby enhancing the emphasis on UAV target region features within the search frame. Furthermore, to enhance the feature extraction capabilities of the network for IRUT without adding extra computational burden, we propose a novel tracking-specific target-aware contextual attention knowledge distiller (TCAKD). It transfers the target prior from the teacher network to the student model, significantly improving the student network's focus on informative regions at each hierarchical level of the backbone network. Extensive experiments on real infrared UAV datasets demonstrate that the proposed approach outperforms state-of-the-art target trackers under complex backgrounds while achieving a real-time tracking speed.
1 January 2026 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing