OCR of Graphic
Comics Novels

Overview

This project leverages advanced image processing techniques to extract text from multiple PDF files and generate customizable tables with a cell structure. The intuitive interface allows users to effortlessly add, delete, or modify cells and their structures, tailored for seamless data organization. The designed border-less cell structure ensures smooth data transfer. Users can export cells to Excel or JSON format for further analysis. The aim is to offer an intuitive solution for efficient data management from PDFs.

Challenges

Language Barrier:
Many comic fans worldwide can’t fully enjoy graphic novels due to language barriers, limiting global accessibility and engagement.
Inconvenient Reading:
Physical comics are impractical during commutes or travels, and existing digital platforms might not offer optimal mobile reading experiences.
Text Extraction Difficulty:
Extracting text from comics is challenging due to irregular placement and artistic elements, hampering features like translation.

Solution

Multilingual Translation:
Our iOS/Android app lets users read graphic novels in their preferred languages. OCR technology translates comic text, overcoming language barriers.
Optimized Mobile Reading:
Our app offers a smooth mobile reading experience, optimized for iOS and Android devices, ensuring comics can be enjoyed on the go.
Accurate OCR:
Employing advanced OCR engines and a deep learning segmentation pipeline, our solution accurately extracts comic text for seamless features like translation and search.

Development Process

1

Research

2

Planning

3

Designing

4

Development

5

Maintenance

Sales & ROI

As a result of the new properly designed website, our client FINews, was able to engage with the
audience well and close more sales in a short span of 2 months. They gained the ROI after 3
months with our assistance.

20%

Conversion rate
in 2022

80%

Increase in monthly
revenue

Team & Role

A dedicated team of 12 individuals contributes their expertise to this project, including:
  • Two Data Science Engineers
  • Data Collection and Cleaning Specialists
  • UI/UX Designers
  • Blockchain Experts
  • Backend and Frontend Engineers

Tools / Technologies

The project leverages a robust technology stack to ensure efficient performance and reliability.

OpenVeno

Python

FASTAPI

PostgreSQL

Conclusion

FINews’s objective was to build a user-friendly and aesthetically pleasing website that would
encourage greater traffic and sales. Deline Media was able to develop a user-friendly and
responsive website for them which enabled their customers to explore the capital market and
current forex rates fast enough.

Technical Achievements

Versatile OCR Expertise:
Proficient usage of diverse OCR engines, such as Tesseract, Paddle OCR, and Easy OCR, to extract text from comics.

Deep Learning Segmentation Pipeline:
Development of a sophisticated deep learning pipeline for segmentation, accurately identifying text regions for subsequent OCR processing.

Let's Connect

New York, USA
Lahore, Pakistan
Right Reserved © Deline Media 2023

Let's Connect