Signal, Image and Video Processing Обработка сигналов

Adaptive weight hybrid order total variation image denoising model based on structure tensor

Bi, Hui, Yang, Ke, Han, Lei, Wang, Zhixiao
Springer Nature · 2026-11-13 · DOI 10.1007/s11760-026-05890-6

Аннотация

In recent years, image denoising methods based on partial differential equations (PDEs) have demonstrated promising application prospects due to their effective denoising performance. However, traditional PDE-based denoising methods generally suffer from problems such as noticeable staircase effects and blurred image edges, making it difficult to achieve ideal denoising results. To address these shortcomings, this paper proposes an adaptive weighted hybrid order total variation image denoising model based on structure tensors. The model constructs an edge detection operator using the eigenvalues of the structure tensor, and achieves ideal denoising effects by fusing two types of regularization terms and adaptively adjusting their weights to adapt to the local structure of the image. Finally, experimental results demonstrate that the proposed method can effectively suppress the staircase effect of the total variation model while accurately preserving image texture details, showing strong competitiveness compared with other image denoising methods.

Читать у издателя →