Deformed text correction method and system and character recognition method

A text and deformity technology, applied in the field of image processing, can solve the problems of difficult to identify irregular shapes and affect the correction effect, and achieve the effect of good correction effect and small amount of calculation.

Pending Publication Date: 2020-11-20
SHANGHAI MININGLAMP ARTIFICIAL INTELLIGENCE GRP CO LTD
View PDF13 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The Hough transform is better for character recognition on specific curves, but it is difficult to recognize irregular shapes
The method of deep learning requires a large number of samples for training. When the number of samples is insufficient, the correction effect will also be affected.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Deformed text correction method and system and character recognition method
  • Deformed text correction method and system and character recognition method
  • Deformed text correction method and system and character recognition method

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment 1

[0057] figure 2 It is a schematic flow chart of the malformed text correction method in the embodiment of the present invention, refer to figure 2 As shown, the malformed text correction method of the embodiment of the present invention includes:

[0058] The document image acquisition step S10 is used to acquire the target document image to be operated;

[0059] The model generation step S20 is used to generate a Gaussian heat map based on the pixel distribution of the document image. The Gaussian heat map is a pixel probability model of the document image, and the probability of occurrence of pixels is visually reflected through the Gaussian heat map;

[0060] The text segmentation step S30 is used to obtain the text boundary of the Gaussian heat map through edge detection, and segment the text of the document image according to the text boundary to obtain multiple text area images corresponding to the text of the document image;

[0061]The text correction step S40 is u...

specific Embodiment 2

[0100] Only the differences between this embodiment and specific embodiments are described below, and the similarities will not be repeated here. Figure 5 It is a schematic flow chart of the malformed text correction method of this embodiment, Figure 6 Shown is a schematic block diagram of the structure of the malformed text correction system of this embodiment, refer to Figure 5-6 As shown, the difference between this embodiment and the specific embodiment is:

[0101] In the deformed text correction method of this embodiment, the text segmentation step S30 further includes: a text area image indexing step S301, which is used to record the relative positional relationship of each text area image through a position index, and the relative positional relationship can be a plurality of text areas The positional relationship between the images may also be the positional relationship between each character area image and the document image. The text correction step S40 also i...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a malformed text correction method and system and a text recognition method. The method comprises a document image obtaining step which is used for obtaining a target document image; a model generation step for generating a Gaussian thermodynamic diagram based on the pixel distribution of the document image, the Gaussian thermodynamic diagram being a pixel probability modelof the document image, and visually reflecting a pixel occurrence probability through the Gaussian thermodynamic diagram; a character segmentation step for obtaining a character boundary of the Gaussian thermodynamic diagram through edge detection, and segmenting the characters of the document image according to the character boundary to obtain a plurality of character region images correspondingto the characters of the document image; and a text correction step for performing matrix transformation on each character region image by using singular value decomposition to obtain a corrected character region image so as to obtain a corrected document image. According to the scheme, a relatively good correction effect on deformed characters arranged in any shape is achieved, and the calculation amount is small.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a deformed text correction method, system and character recognition method. Background technique [0002] For distorted lines of text, such as text in circular signs, curved documents, curved documents, etc., such as figure 1 Malformed text as shown in . At present, the correction of deformed text mainly uses the method of Hough transform or deep learning. The Hough transform is better for character recognition of specific curves, but it is difficult to recognize irregular shapes. The method of deep learning requires a large number of samples for training, and when the number of samples is insufficient, the correction effect will also be affected. Contents of the invention [0003] In order to solve the above-mentioned technical problems, the present invention proposes a deformed text correction method, system, and character recognition method, which can...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06K9/32G06K9/34
CPCG06V10/243G06V20/62G06V30/153
Inventor 安达
Owner SHANGHAI MININGLAMP ARTIFICIAL INTELLIGENCE GRP CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap