Text detection method, device, electronic device and computer storage medium

A text detection and text technology, applied in the computer field, can solve problems such as large amount of calculation, time-consuming, and large consumption of computing resources, and achieve the effects of improving efficiency and speed, reducing calculation amount, and saving computing resources

Active Publication Date: 2021-02-26
BEIJING YIZHEN XUESI EDUCATION TECH CO LTD
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, this method requires a large amount of calculation, which not only consumes a lot of computing resources, but also takes a long time, which reduces the speed of text detection

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
  • Text detection method, device, electronic device and computer storage medium
  • Text detection method, device, electronic device and computer storage medium
  • Text detection method, device, electronic device and computer storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0028] Embodiment 1 of the present application provides a text detection method, such as figure 1 as shown, figure 1 It is a flow chart of a text detection method provided in the embodiment of the present application, and the text detection method includes the following steps:

[0029] Step S101 , perform feature extraction on the text image to be detected, and obtain a text region probability map and a text region number probability map corresponding to the text image to be detected.

[0030] It should be noted that the text detection method in the embodiment of the present application is applicable to text detection with various text densities, including but not limited to regular density text, dense density text, sparse density text, especially dense density text. Among them, specific indicators for determining whether a certain text is a dense text can be appropriately set by those skilled in the art according to the actual situation, including but not limited to: accordi...

Embodiment 2

[0043] Embodiment 2 of the present application is based on the solution of Embodiment 1. Optionally, in an embodiment of the present application, step S102 may be implemented as step 102a and step 102b.

[0044] Exemplarily, in step 102a, binarize the text region probability map to obtain a text region binary map; in step 102b, perform an AND operation on the text region binary map and the text region number probability map to obtain a text region number map.

[0045] Wherein, through the AND operation, valid pixels in the text area number probability map can be retained, or noise pixels in the text area number probability map can be filtered out.

[0046] Optionally, in one embodiment of the present application, step 102b is implemented in the following manner: determining the pixel point corresponding to the pixel point representing the text in the text region number probability map and the text region binary map as an effective pixel point, the text area number probability ...

Embodiment 3

[0059] Embodiment 3 of the present application is based on the solutions of Embodiment 1 and Embodiment 2, wherein step S101 can also be implemented as the following steps 101a-101d.

[0060] Step 101a, performing first text feature extraction on the text image to be detected.

[0061] In the embodiment of the present application, when performing feature extraction on the text image to be detected, the text image to be detected can be input into the residual network part (such as the Resnet network) to extract the first text feature, such as extracting texture, edge, and corner points from the input image and semantic information, which are represented by 4 sets of feature maps of different sizes. Take the text image to be detected as the original image, and the Resnet network extracts the features of the original image as an example. The Resnet18 network is constructed by connecting four blocks in series. Each block includes several layers of convolution operations. The outpu...

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

Embodiments of the present application provide a text detection method, device, electronic equipment, and computer storage medium. The text detection method includes: first performing feature extraction on the text image to be detected, and obtaining the text region probability map and text region corresponding to the text image to be detected. numbering probability map; and then determine the text area numbering map according to the text area binary map corresponding to the text area probability map and the text area numbering probability map; according to the text area numbering map and the different numbering thresholds corresponding to the text lines in different blocks, you can The coordinates of each real text region in the text image to be detected are obtained, and the text detection result of the text image to be detected is obtained. Through the solution of the embodiment of the present application, there is no need for repeated convolution, and it is not necessary to traverse the feature map and classify whether there is text in the frame one by one, which reduces the amount of text detection calculations, saves computing resources, and improves text detection. Efficiency and speed.

Description

technical field [0001] The embodiments of the present application relate to the field of computer technologies, and in particular to a text detection method, device, electronic equipment, and computer storage medium. Background technique [0002] Text detection is a technology that detects text regions in images and marks their bounding boxes. Text detection has a wide range of applications and is a pre-step for many computer vision tasks, such as image search, text recognition, identity authentication, and visual navigation. [0003] The main purpose of text detection is to locate the position of a text line or character in an image. Currently, a popular text detection method is a text detection method based on a sliding window. Based on the idea of ​​general target detection, this method sets a large number of anchor boxes with different aspect ratios and different sizes, and uses these anchor boxes as sliding windows to perform convolution operations on the image or the f...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/62G06K9/32G06K9/46G06N3/04G06N3/08
CPCG06N3/08G06V10/25G06V10/40G06N3/045G06F18/2415
Inventor 刘军秦勇
Owner BEIJING YIZHEN XUESI EDUCATION TECH 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