Kallind / Text_Detector

This project is primarily a text detector for 16 Segment Display.

Text_Detector

Text extraction from a 16-segment display using computer vision techniques.

Table of Contents

1. Introduction
2. Setup
3. Usage
4. Scripts Overview
5. Web UI
6. Contributing

Introduction

Text_Detector is designed to detect and extract text from videos and images, especially from a 16-segment display. The project utilizes OpenCV for image and video processing and EasyOCR for text recognition. A Flask-based UI is also provided for visualization.

Setup

1. Make sure Python 3.x is installed.
2. Clone the repository:

git clone https://github.com/Dv04/Text_Detector
cd Text_Detector

3. Install the required libraries:

pip install -r requirements.txt

Usage

1. Extract frames from video:

python ComputerVision/CameraVision.py

2. Detect text from video:

python ComputerVision/ComputerVision.py

3. Detect and test text extraction from image:

python ComputerVision/ImageVision.py

4. Launch the Flask UI:

python ui/app.py

Then access the UI at http://127.0.0.1:5000/.

Scripts Overview

1. CameraVision.py: Extracts frames from videos, specifically every 21st frame.
2. ComputerVision.py: Processes videos to extract text, leaning on ImageVision.py for detecting text inside red boxes.
3. ImageVision.py: Discovers red boxes in images, extracting text from these zones using EasyOCR.
4. app.py: The Flask UI, showcasing extracted frames and detected text.

Web UI

The web UI is built using Flask, allowing users to visualize text detection results:

• index.html: Main UI for viewing images and extracted text.
• style.css: Contains styling for the UI.

For a detailed walkthrough of the UI components, view the code in the ui directory.

Contributing

Before contributing, please review the guidelines in CONTRIBUTING.md. Adherence to these rules ensures smooth collaboration and code integration.

---