Jove
Visualize
Contact Us

Related Experiment Video

Updated: Nov 17, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

792

Text Detection Using Multi-Stage Region Proposal Network Sensitive to Text Scale.

Yoshito Nagaoka1, Tomo Miyazaki1, Yoshihiro Sugaya1

  • 1Graduate School of Engineering, Tohoku University, Sendai 9808579, Japan.

Sensors (Basel, Switzerland)
|February 12, 2021
PubMed
Summary

This study introduces a new convolutional neural network (CNN) for intelligent sensors to improve small text detection. The novel architecture enhances feature extraction across multiple resolutions, crucial for accurate text recognition.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

TAMC: Textual Alignment and Masked Consistency for Open-Vocabulary 3D Scene Understanding.

Sensors (Basel, Switzerland)·2024
Same author

Machine learning-based identification of the risk factors for postoperative nausea and vomiting in adults.

PloS one·2024
Same author

JPEG Image Enhancement with Pre-Processing of Color Reduction and Smoothing.

Sensors (Basel, Switzerland)·2023
Same author

Spectrum Correction Using Modeled Panchromatic Image for Pansharpening.

Journal of imaging·2021
Same author

Graph Neural Networks with Multiple Feature Extraction Paths for Chemical Property Estimation.

Molecules (Basel, Switzerland)·2021
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Area of Science:

  • Computer Vision
  • Artificial Intelligence
  • Sensor Technology

Background:

  • Intelligent sensors increasingly utilize text detection capabilities.
  • Detecting small text presents a significant challenge in current systems.
  • Existing methods struggle with scale variations in text recognition.

Purpose of the Study:

  • To propose a novel convolutional neural network (CNN) architecture for text detection.
  • To enhance the sensitivity of text detection models to varying text scales.
  • To address the limitations of current methods in small text detection.

Main Methods:

  • Extraction of multi-resolution feature maps using multi-stage convolution layers.
  • Maintaining feature size and preventing information loss during extraction.
Keywords:
convolutional neural networksmultiple scalesscene text detection

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

650
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.9K

Related Experiment Videos

Last Updated: Nov 17, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

792
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

650
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
05:56

Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application

Published on: April 14, 2023

2.9K
  • Designing the CNN architecture with consideration for receptive field size for proposal generation.

    • Main Results:

    • The proposed CNN architecture demonstrates improved performance in text detection.
    • Multi-resolution feature extraction effectively preserves information for small text.
    • Experimental validation highlights the critical role of receptive field size in performance.

    Conclusions:

    • The novel CNN architecture is effective for text detection, particularly for small text.
    • Multi-resolution feature maps and controlled receptive fields are key to the model's success.
    • This research contributes to advancing intelligent sensor capabilities in text recognition.