基于张量双低秩约束的遥感图像去噪方法

Remote Sensing Image Denoising Method Based on Tensor Double Low Rank Constraint

传统的张量去噪方法往往只使用Tucker秩或CP秩进行低秩约束。事实上,它们代表了不同的高维数据结构。为了进一步提升遥感图像去噪性能,受张量补全方法的启发,本文提出基于张量双低秩约束的遥感图像去噪方法。该方法在低秩张量近似中通过联合最小化CP秩和Tucker秩来同时利用这两种数据结构进行去噪。仿真实验结果表明,本方法在评价指标和视觉效果上均优于对比方法。

Traditional tensor denoising methods tend to use only Tucker rank or CP rank for low rank constraints.Indeed,they represent different high-dimensional data structures.To further improve the denoising performance of remote sensing images,inspired by the tensor completion method;this paper proposes a remote sensing image denoising method based on the tensor double low-rank constraint.This method simultaneously utilizes both data structures to denoise by jointly minimizing the CP rank and Tucker rank in the low-rank tensor approximation.Simulation experiment results show that this method is better than the comparison methods in both evaluation index and visual effects.