Abstract

According to one implementation, a system for performing re-noising and neural network (NN) based image enhancement includes a computing platform having a processing hardware and a system memory storing a software code, a noise synthesizer, and an image restoration NN. The processing hardware is configured to execute the software code to receive a denoised image component and a noise component extracted from a degraded image, to generate, using the noise synthesizer and the noise component, synthesized noise corresponding to the noise component, and to interpolate, using the noise component and the synthesized noise, an output image noise. The processing hardware is further configured to execute the software code to enhance, using the image restoration NN, the denoised image component to provide an output image component, and to re-noise the output image component, using the output image noise, to produce an enhanced output image corresponding to the degraded image.

Background

BACKGROUND (1) Image processing techniques are important to many applications. For example, in image enhancement, a degraded version of an image is processed using one or more techniques to de-blur or sharpen the degraded image, or to reduce noise present in the degraded image. Despite its usefulness for recovering degraded image features, image enhancement can present significant challenges. For example, legacy video content for which remastering may be desirable is often available only in interlaced, noisy, and low resolution formats. As a result the remastering process has to be carefully engineered to enhance desirable image features while avoiding exaggeration of degraded features in the image. Nevertheless, in the interests of preserving the original aesthetic of legacy content undergoing remastering, as well as to respect the artistic intent of its creators, it may be advantageous or desirable to retain some nominally degraded features, such as some noise, for example, to enhance the apparent authenticity of remastered content, as well as the imagery it contains.

Claims

1. A system comprising: a computing platform including a processing hardware and a system memory storing a software code, a noise synthesizer, and an image restoration neural network (NN): the processing hardware configured to execute the software code to: receive a denoised image component and a noise component extracted from a degraded image; generate, using the noise synthesizer and the noise component, a synthesized noise corresponding to the noise component; interpolate, using the noise component and the synthesized noise, an output image noise; enhance, using the image restoration NN, the denoised image component to provide an output image component; and re-noise the output image component, using the output image noise, to produce an enhanced output image corresponding to the degraded image. || 8. A system comprising: a computing platform including a processing hardware and a system memory storing a software code, a noise synthesizer, a signal splitting neural network (NN), and an image restoration NN: the processing hardware configured to execute the software code to: receive a degraded input image; extract, using the signal splitting NN, a denoised image component and a noise component from the degraded input image; generate, using the noise synthesizer and the noise component, a synthesized noise corresponding to the noise component; interpolate, using the noise component and the synthesized noise, an output image noise; enhance, using the image restoration NN, the denoised image component to provide an output image component; and re-noise the output image component, using the output image noise, to produce an enhanced output image corresponding to the degraded input image. || 15. A system comprising: a computing platform including a processing hardware and a system memory storing a software code, a noise synthesizer, a de-interlacer neural network (NN), a signal splitting NN, and an image restoration NN: the processing hardware configured to execute the software code to: receive a degraded interlaced input image; de-interlace, using the de-interlacer NN, the degraded interlaced input image to provide a degraded input image; extract, using the signal splitting NN, a denoised image component and a noise component from the degraded input image; generate, using the noise synthesizer and the noise component, a synthesized noise corresponding to the noise component; interpolate, using the noise component and the synthesized noise, an output image noise; enhance, using the image restoration NN, the denoised image component to provide an output image component; and re-noise the output image component, using the output image noise, to produce an enhanced output image corresponding to the degraded input image.

Re-noising and neural network based image enhancement

Abstract

Background

Claims