Image correction method and device, and computer readable storage medium

Abstract

The invention discloses an image correction method and device, and a computer readable storage medium, and belongs to the technical field of image processing. The image correction method comprises thesteps that aiming at one or more of a plurality of correction steps involved in image correction, one or more neural network models are constructed correspondingly; based on the multiple correction steps and the one or more neural network models, process configuration files are generated; the flow logic between the multiple correction steps is configured in the process configuration files, and animage correction process is generated; and when image correction instructions are received, target images are corrected according to the image correction process, and the corrected target images areoutput. According to the image correction method and device, and the computer readable storage medium, image information with text data under different scenes can be corrected into normative image information easy to detect and recognize, and the neural network models are constructed in a deep learning mode and applied to the correction steps of the image correction process, so that the higher accuracy rate for image correction is obtained.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to an image correction method, device and computer-readable storage medium. Background technique [0002] Optical Character Recognition (OCR) is usually used in document processing and recognition. It converts the text of various bills, newspapers, books, manuscripts and other printed materials into image information through optical input methods such as scanning and photographing, and then uses Text recognition technology converts image information into usable computer input technology. Optical character recognition technology has always been an important technical means to assist people in image recognition, document reading, analysis and processing, and is widely used in banking, insurance, taxation, auditing, legal and other industries. [0003] The use of optical character recognition for document images has certain requirements for document images. If the...

AI Technical Summary

Problems solved by technology

[0004] There are many defects in the traditional image correction method. For example, in terms of process, it lacks a general and systematic process, and it is difficult to correct image information with text data in real scenes into standardized image information that is easy to detect and identify; For another example, in terms of algorithms, it lacks versatility and accuracy, and cannot cover noisy and distorted images in all scenarios. Some of the algorithm units, such as perspective transformation correction, watermark removal, and stamp removal, are difficult to meet expectations. Effect

Images