https://github.com/burakahmet/instant-handwritten-digit-detection

Instant handwritten digit detection with OpenCV and TensorFlow.
https://github.com/burakahmet/instant-handwritten-digit-detection

Last synced: 19 days ago
JSON representation

Instant handwritten digit detection with OpenCV and TensorFlow.

• Host: GitHub
• URL: https://github.com/burakahmet/instant-handwritten-digit-detection
• Owner: BurakAhmet
• License: apache-2.0
• Created: 2024-03-20T08:32:13.000Z (9 months ago)
• Default Branch: main
• Last Pushed: 2024-03-20T12:28:28.000Z (9 months ago)
• Last Synced: 2024-03-21T09:49:09.827Z (9 months ago)
• Language: Jupyter Notebook
• Size: 45.4 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Handwritten Digit Detection

This project implements a handwritten digit detection system using OpenCV and TensorFlow's convolutional neural network (CNN) model. The goal of the project is to allow users to draw digits on a canvas, after which the program detects and recognizes the digits, displaying the result alongside the original drawing. 

## Preview

https://github.com/BurakAhmet/Instant-Handwritten-Digit-Detection/assets/89780902/571a6ae5-f02f-4dc8-bac0-5a0b571aa0e9

## Model

**Final validation loss: 0.0144**

**Final validation accuracy: 0.9960**

For more details you can check [model_training.ipynb](https://github.com/BurakAhmet/Instant-Handwritten-Digit-Detection/blob/main/model_training.ipynb) file.

## Usage

1- Run the program.

2- Draw digits on the canvas using the left mouse button.

3- Release the mouse button to signal drawing completion.

4- View the recognized digits alongside the original drawing.

5- Press "q" to quit the program or "r" to clear the canvas.

## Technologies Used

* Python: The project is developed using Python programming language.

* TensorFlow: Used for Creating the model, training the data and making predictions.

* OpenCV: Used for image preprocessing, visualization of results, user interface.

* NumPy: Employed for array manipulation and normalization of input data.

* Google Colab: Used for fast model training with GPUs.