n Image Noise Level Estimation Based on Tensor T-Product

he image noise level indicates the degree to which an image is contaminated bynoise, and serving as a crucial parameter in the image processing process. For the third-order tensor corresponding to a color image, traditional methods for estimatingthe noise level will disrupt the data structure of the tensor. In order to solve the aforementioned problems, we directly select the tensor using a sliding block of size  and then re-arrange it. The main innovations of this paper are as follows:        (1) A new noise level estimation model based on tensor decomposition is proposed. The third-order tensor is decomposed into the form of a covariance matrix according to the definition of the T-product, and then the noise level of the colorimage is estimated, and good experimental results were achieved in the experiment.       (2) Through theoretical analysis, it has been proved that the multiple eigenvalues of the matrix obtained by T-product of the third-order tensor have a direct relationship with the noise level of the color image.