Image super-resolution and coloring method and system and electronic equipment

Abstract

The invention relates to an image super-resolution and coloring method, an image super-resolution and coloring system and electronic equipment. The method comprises the following steps: step a, designing a new network model by combining image super-resolution and coloring, wherein the network model comprises an encoding module and a decoding module; b, inputting the low-resolution grayscale imageinto a network model; and step c, extracting the semantic feature map of the low-resolution grayscale image through the encoding module, transmitting the semantic feature map to a decoding module, superposing the semantic feature maps of all levels by the decoding module, and outputting a high-resolution color map. The new network model is designed for the super-resolution and coloring problems, the super-resolution and coloring tasks are processed together, the low-resolution grey-scale map is directly mapped to the high-resolution color map, the visual impression of people is improved, and resources and time are saved.

Description

technical field [0001] The present application belongs to the technical field of image processing, and in particular relates to an image super-resolution and coloring method, system and electronic equipment. Background technique [0002] In the field of image processing, image super-resolution and colorization have been two hot research areas. Image super-resolution technology refers to recovering a high-resolution image from a low-resolution image or image sequence. This technology was first proposed in the field of optics, and is now widely used in the fields of image compression, medical imaging, and remote sensing imaging. Traditional super-resolution techniques include: interpolation-based methods (such as: nearest neighbor interpolation method, bilinear interpolation method, bicubic interpolation method, etc.), reconstruction-based methods (such as: non-uniform interpolation method, iterative back-projection method, maximum Posterior probability method, etc.), etc. ...

AI Technical Summary

Problems solved by technology

[0005] To sum up, in the existing technology, for some tasks that require both super-resolution and coloring of images (such as: processing some old photos and videos), it is necessary to separately build image super-resolution and coloring. model, train two network models, and process these two tasks separately, which is a waste of resources and time

At the same time, regardless of super-resolution first and then coloring, or first coloring and then super-resolution, due to the error produced by the previous task, it is very difficult for the latter task, which is not conducive to recovering high-resolution color images.

Images