Image distortion coefficient extraction method, distortion correction method and system, and electronic equipment

Abstract

The invention discloses an image distortion coefficient extraction method, a distortion correction method and system and electronic equipment. The image distortion coefficient extraction method comprises the following steps: extracting a distortion coefficient on a to-be-corrected image by adopting a pre-trained artificial intelligence model, wherein the artificial intelligence model is a U-Net network. The image distortion correction method comprises the following steps: extracting a distortion coefficient of an image to be corrected according to an image distortion coefficient extraction method; and performing distortion correction on the to-be-corrected image based on the distortion coefficient. The distortion detection method is simple and rapid, the distortion degree in the image can be rapidly and accurately distinguished through a simple model trained by machine learning, and the degree of pixel-by-pixel can be achieved; and according to the distortion correction method, an advanced deep learning algorithm is used, so that the robustness is high, the consumed time is short, and the effect is very good.

Description

technical field [0001] The invention belongs to the technical field of machine learning, and in particular relates to an image distortion coefficient extraction method, a distortion correction method and system, and an electronic device and a computer-readable medium using the same. Background technique [0002] Whether it is in daily life or in the business of Internet companies, when you need to process document images, you often encounter the problem of distortion of key elements such as text and graphics in document images. Once the key elements are distorted, it will affect various subsequent applications. cause difficulty. If the distorted image can be reconstructed into a flat image as much as possible, it will be of great benefit to both user experience and subsequent business development. The traditional distortion correction method, for example, can divide the text recognized from the captured image into multiple connected domains by means of text domain segmentat...

AI Technical Summary

Problems solved by technology

This method works well in large-scale text document scenes, but it is difficult to achieve good results in the case of documents with mixed graphics and text.

In recent years, a number of adaptive methods using deep learning have also emerged. Traditional methods and early deep learning methods often have poor results, poor robustness, and take a long time to meet the needs of mass use.

Images