UTRNet: High-Resolution Urdu Text Recognition in Printed Documents

Rahman, Abdur; Ghosh, Arjun; Arora, Chetan

doi:10.1007/978-3-031-41734-4_19

Cited by 3 publications

(1 citation statement)

References 68 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A noticeable research gap exists for cursive languages like Kannada, Hindi, and Tamil, especially in natural show promise in recognizing cursive texts [18]- [20], the challenge is amplified with texts in various natural scenes due to their complexities and variability in backgrounds, fonts, sizes, and colors. Furthermore, much of the existing literature is confined to studying isolated characters and scripts [21]- [23].…”

Section: Introductionmentioning

confidence: 99%

CRNN model for text detection and classification from natural scenes

Prakash,

Yeliyur Hanumanthaiah,

Bannur Mayigowda

2024

IJ-AI

View full text Add to dashboard Cite

<span lang="EN-US">In the emerging field of computer vision, text recognition in natural settings remains a significant challenge due to variables like font, text size, and background complexity. This study introduces a method focusing on the automatic detection and classification of cursive text in multiple languages: English, Hindi, Tamil, and Kannada using a deep convolutional recurrent neural network (CRNN). The architecture combines convolutional neural networks (CNN) and long short-term memory (LSTM) networks for effective spatial and temporal learning. We employed pre-trained CNN models like VGG-16 and ResNet-18 for feature extraction and evaluated their performance. The method outperformed existing techniques, achieving an accuracy of 95.0%, 96.3%, and 96.2% on ICDAR 2015, ICDAR 2017, and a custom dataset (PDT2023), respectively. The findings not only push the boundaries of text detection technology but also offer promising prospects for practical applications.</span>

show abstract

Section: Introductionmentioning

confidence: 99%

CRNN model for text detection and classification from natural scenes

Prakash,

Yeliyur Hanumanthaiah,

Bannur Mayigowda

2024

IJ-AI

View full text Add to dashboard Cite

show abstract

MULDT: Multilingual Ultra-Lightweight Document Text Detection for Embedded Devices

Gayer,

Arlazarov

2024

IEEE Access

View full text Add to dashboard Cite

Urdu Cursive Word Recognition Using an Advanced Intelligent Model of Optimized Deep Learning

Asha,

Uma,

Kavin

et al. 2024

Advanced Intelligence Systems and Innovation in Entrepreneurship

View full text Add to dashboard Cite

Many historically significant documents are only accessible in paper record form, making text recognition a crucial problem in the arena of digital image processing. Text recognition techniques primarily aim to convert paper documents into digital files that can be easily managed in a database or other server-based entity. Size, colour, font, orientation, backdrop complexity, occlusion, illumination, and lighting all make text identification more difficult in photos from real-world settings. Variations in writing style, several forms of the same letter, linked text, ligature diagonal, and condensed text make Urdu text identification more difficult than with non-cursive scripts. To separate the spatial correlation and appearance correlation (DSSAC) of the mapped convolutional channel, the suggested intelligent model employs the deep separable convolutional layers in place of the conventional design in the U-Net. To achieve cursive region, capture, the research offers a model called DSSAC-RSC.

show abstract

UTRNet: High-Resolution Urdu Text Recognition in Printed Documents

Cited by 3 publications

References 68 publications

CRNN model for text detection and classification from natural scenes

CRNN model for text detection and classification from natural scenes

MULDT: Multilingual Ultra-Lightweight Document Text Detection for Embedded Devices

Urdu Cursive Word Recognition Using an Advanced Intelligent Model of Optimized Deep Learning

Contact Info

Product

Resources

About