Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Enhancing camera-captured Devanagari documents via geometric filtering for improved vision-language model text

Anup Kelkar¹, Parag Deshpande², O G Kakde²

¹Department of Computer Science and Engineering, Indian Institute of Information Technology (IIIT), Nagpur, Maharashtra, 440024, India.

|June 4, 2026

Summary

Related Concept Videos

Improving Translational Accuracy

Improving Translational Accuracy

Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...

Improving Translational Accuracy

Improving Translational Accuracy

Base complementarity between the three base pairs of mRNA codon and the tRNA anticodon is not a failsafe mechanism. Inaccuracies can range from a single mismatch to no correct base pairing at all. The free energy difference between the correct and nearly correct base pairs can be as small as 3 kcal/ mol. With complementarity being the only proofreading step, the estimated error frequency would be one wrong amino acid in every 100 amino acids incorporated. However, error frequencies observed in...

You might also read

Related Articles

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

Sort by

Same author

Performance optimisation of deep learning models using majority voting algorithm for brain tumour classification.

Computers in biology and medicine·2021

Same author

Enhanced bioavailability and bioefficacy of an amorphous solid dispersion of curcumin.

Food chemistry·2014

Same author

Development and validation of LC-MS/MS method for the estimation of β-hydroxy-β-methylbutyrate in rat plasma and its application to pharmacokinetic studies.

Biomedical chromatography : BMC·2012

See all related articles

This study introduces an image enhancement method for Devanagari text recognition, improving text extraction accuracy in Vision-Language Models (VLMs) for non-English languages.

Area of Science:

Computer Science
Artificial Intelligence
Natural Language Processing

Background:

Modern Vision-Language Models (VLMs) struggle with text extraction in non-English languages, especially Devanagari script, due to data scarcity.
Recognizing Devanagari text in diverse materials presents significant accuracy and efficiency challenges for existing platforms.
Effective text extraction requires pre-processing to maintain crucial information and linguistic integrity.

Purpose of the Study:

To propose an advanced image enhancement technique specifically designed for Devanagari text.
To improve the accuracy and efficiency of text extraction for Devanagari script within VLM platforms.
To address the limitations of current VLM capabilities in processing regional languages.

Main Methods:

Keywords:

Camera-captured document image Devanagari character’s geometric properties Text extraction VLM platforms

Related Experiment Videos

Developed an image enhancement method utilizing geometric filters and clustering algorithms tailored for Devanagari characters.
Leveraged the unique geometrical properties of Devanagari script to enhance sentence detection.
The technique was designed to be robust against variations in font size, orientation, document rotation, size, and color.

Main Results:

The proposed method accurately detects Devanagari sentences even in degraded, hand-held camera-captured images.
Demonstrated superior performance compared to existing Optical Character Recognition (OCR) pre-processing techniques.
Achieved a 95.23% success rate in locating Devanagari sentences, significantly enhancing VLM text extraction accuracy.

Conclusions:

The novel image enhancement approach effectively overcomes challenges in Devanagari text recognition.
This method substantially improves the performance of Vision-Language Models for text extraction in Devanagari script.
The study highlights a viable solution for expanding VLM applicability to under-resourced languages.