dv04 / text_detector

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

Hi there 👋

Nice to meet you. Not much but I try to do something and everything I can.

👨🏻‍💻: About Me :

I am a Student Intern Open Source Developer from Gujarat, India.

• 🤖: I work as an ML Intern and contribute to the making of different types of Machine learning models and AIs.

• 🌱 Currently Exploring Deep Learning and Reinforcement Learning.

• ⚡ My Free time goes in doing competitive programming in Leetcode, Codechef and Codeforeces.

• 📁 Visit my main Repository

• 📫Reach me at:

💻: Languages and Tools :

                                       

🤝🏻: Connect with Me :

📈: My Github Stats :

(Create with github-readme-stats)

text_detector's Issues

Code Refactoring

Description:

Modularize the code to improve readability, maintainability, and scalability.

Tasks:

• Break down monolithic code files into smaller, manageable modules.
• Ensure functions and classes adhere to the Single Responsibility Principle.
• Update function and variable names to be descriptive and adhere to naming conventions. (wherever needed)
• Organize the directory structure for better navigation and understanding.
• Ensure that there is a clear and consistent coding style throughout the project.
• Add comments and documentation where necessary to explain complex or non-intuitive code.

Note:

• A PR can be generated if at least one task from above is completed

Model Evaluation

Description:

Implement metrics to evaluate the performance and accuracy of the CNN model used for detecting and predicting 16-segment displays.

Tasks:

• Implement confusion matrix, precision, recall, and F1-score to evaluate model performance.
• Compare the model's performance on a validation set.
• Visualize the evaluation metrics in a comprehensible manner, e.g., plots or dashboards.
• Document the evaluation process and results.

Create Comprehensive README.md

Description:

The repository lacks a detailed README.md file. A well-structured and informative README is crucial for understanding the project, its structure, and how to use or contribute to it.

Tasks:

• Provide a project description explaining the purpose and functionality of Text_Detector.
• Include setup instructions detailing how to get the project running.
• Document usage examples with code snippets and screenshots if possible.
• List dependencies and provide information on how to install them.
• Add a section for contributors on how to report issues, propose changes or contribute to the project.
• Include a license and citation information if applicable.

Text Detection

Description:

Develop and implement a text detection module capable of identifying the 16-segment displays in the given images.

Tasks:

• Research and choose a suitable text detection method.
• Implement the text detection module.
• Test the module with a subset of images.
• Document the method, implementation, and any external libraries used in the repository.

Image Preprocessing

Description:

Implement image preprocessing to enhance the quality and ensure consistent input for the CNN model. This includes resizing, thresholding, and potentially augmenting the images to increase the dataset size.

Tasks:

• Implement resizing to ensure consistent image dimensions.
• Apply thresholding to enhance text visibility.
• Explore image augmentation techniques to expand the dataset.
• Document the preprocessing steps and any external libraries used in the repository.

CSS Enhancements

Description:

Enhance the appearance and responsiveness of the website by updating the CSS styling.

Tasks:

• Improve layout, color scheme, and typography to ensure a visually appealing design.
• Optimize the CSS for different devices and screen sizes to ensure a responsive design.
• Minimize the use of external libraries to reduce load times.
• Ensure consistency in styling across all pages.
• Organize the CSS code and remove any redundant or outdated styles.

Website Integration

Description:

Integrate the Text Detection and Extraction modules into a web application to provide a user interface for interacting with the system.
Design and implement additional pages for the web application to enhance user experience and provide more information.

Tasks:

• Create routes and views for uploading and viewing images.
• Integrate the text detection and extraction modules to process uploaded images.
• Display the detected and extracted text alongside the corresponding images.
• Ensure the website is functioning correctly and debug any issues that arise.
• Document the integration process in the repository.
• Create an about page explaining the purpose and functionality of the application.
• Design a contact page for users to get in touch.
• Implement a help page to guide users on how to use the application.
• Ensure consistency in design across all pages.
• Update navigation to include links to the new pages.

Note:

• Any number of completed Tasks from above would be acceptable for a PR

Text Extraction

Description:

Create a text extraction module to extract and output the recognized text from the detected segments of the images.

Tasks:

• Develop a method to extract the text from the identified segments.
• Implement the text extraction module.
• Test the module on a subset of images to ensure accuracy.
• Document the method, implementation, and any external libraries utilized in the repository.

Dataset Finding & Labeling for 16-Segment Display

Description:

There's a need to obtain a robust dataset of 16-segment display images for training the CNN model. Labeling the dataset accurately is crucial for model performance.

Tasks:

• Search and gather a dataset of 16-segment display images.
• Label the dataset with the correct values displayed.
• Organize the dataset in a manner suitable for training and testing.
• Document the source of the dataset and the labeling process in the repository.