PROJECT TITLE :

External Prior Guided Internal Prior Learning for Real-World Noisy Image Denoising - 2018

ABSTRACT:

Most of existing image denoising ways learn image priors from either an external data or the noisy image itself to get rid of noise. However, priors learned from an external information might not be adaptive to the image to be denoised, while priors learned from the given noisy image might not be correct because of the interference of corrupted noise. Meanwhile, the noise in real-world noisy pictures is terribly advanced, which is tough to be described by simple distributions like Gaussian distribution, making real-world noisy image denoising a very difficult downside. We have a tendency to propose to exploit the information in each external knowledge and therefore the given noisy image, and develop an external previous guided internal previous learning technique for real-world noisy image denoising. We have a tendency to first learn external priors from an freelance set of clean natural images. With the help of learned external priors, we then learn internal priors from the given noisy image to refine the prior model. The external and internal priors are formulated as a set of orthogonal dictionaries to efficiently reconstruct the desired image. In depth experiments are performed on many real-world noisy image datasets. The proposed methodology demonstrates highly competitive denoising performance, outperforming state-of-the-art denoising ways together with those designed for real-world noisy images.