Supercharge Your Innovation With Domain-Expert AI Agents!

Text detection model training method and device, text detection method and device, and equipment

A technology for text detection and model training, applied in neural learning methods, biological neural network models, character and pattern recognition, etc., can solve the problem of low efficiency of scene text

Pending Publication Date: 2021-09-07
上海眼控科技股份有限公司
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] However, texts applied in natural scenes, also known as scene texts, have larger changes in scale, aspect ratio, and especially direction than general objects, making it difficult to detect scene texts using a detection network without preset borders. inefficiency in 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 model training method and device, text detection method and device, and equipment
  • Text detection model training method and device, text detection method and device, and equipment
  • Text detection model training method and device, text detection method and device, and equipment

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0065] figure 1 It is a schematic flow chart of a text detection model training method provided in Embodiment 1 of the present invention. This method is applicable to the situation of improving the efficiency of scene text detection. The method can be executed by a text detection model training device, wherein the device can be controlled by software and / or implemented by hardware, and generally integrated on a terminal device. In this embodiment, the terminal device includes, but is not limited to: computers, personal digital assistants, and other devices.

[0066] Due to the particularity of the aspect ratio of the text itself, the general target detection algorithm is not ideal for detecting text. Therefore, many scholars have borrowed the ideas of target detection to improve Faster R-CNN (Towards Real-Time Object Detection with Region Proposal Networks) and SSD (Single Shot MultiBox Detector). However, Faster R-CNN and SSD have been proved to be effective for text detectio...

Embodiment 2

[0106] figure 2 It is a schematic flowchart of a text detection method provided by Embodiment 2 of the present invention. The method can be executed by a text detection device, wherein the device can be implemented by software and / or hardware, and is generally integrated on a terminal device.

[0107] Such as figure 2 As shown, Embodiment 2 of the present invention provides a text detection method, including the following steps:

[0108] S210. Acquire an image to be tested.

[0109] The image to be tested can be considered as an image to be tested for text in the application stage of the text detection model. The image to be tested and the image sample can be images of the same type. For example, what is recognized is the scene text. The scene text can be considered as the text in the natural scene.

[0110] S220. Input the image to be tested into a text detection model to obtain a center point of a sub-image included in the image to be tested and a boundary point corre...

Embodiment 3

[0144] Figure 9 A schematic structural diagram of a text detection model training device provided by Embodiment 3 of the present invention, which is applicable to improving the efficiency of scene text detection, wherein the device can be implemented by software and / or hardware, and is generally integrated on the terminal device .

[0145] like Figure 9 As shown, the device includes:

[0146] An acquisition module 91, configured to acquire a set number of image samples;

[0147] A segmentation module 92, configured to segment the text regions included in each of the image samples to obtain training sample data;

[0148] The training module 93 is configured to use the training sample data to train the pre-built multi-anchor-free region candidate network to obtain a trained text detection model.

[0149] In this embodiment, the device first acquires a set number of image samples through the acquisition module 91; then through the segmentation module 92 to segment the text ...

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 discloses a text detection model training method and device, a text detection method and device, and equipment. The method comprises the following steps: acquiring a set number of image samples; segmenting a text region included in each image sample to obtain training sample data; and using the training sample data to train a pre-constructed multi-anchor-free region candidate network to obtain a trained text detection model. By utilizing the method, the scene text detection efficiency can be improved.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of culture detection, and in particular to a text detection model training method, a text detection method, device and equipment. Background technique [0002] Text detection is all about locating regions of text in an image and then labeling words or lines of text, usually in the form of bounding boxes. [0003] Due to the particularity of the aspect ratio of the text itself, the general target detection algorithm is not ideal for detecting text. When performing text detection in the related art, a detection network without a preset frame is used. [0004] However, texts applied in natural scenes, also known as scene texts, have larger changes in scale, aspect ratio, and especially direction than general objects, making it difficult to detect scene texts using a detection network without preset borders. The detection efficiency is low. Contents of the invention [0005] Embodi...

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 Applications(China)
IPC IPC(8): G06K9/20G06K9/46G06N3/04G06N3/08
CPCG06N3/04G06N3/08
Inventor 高凯珺
Owner 上海眼控科技股份有限公司
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com