Scrabble Assistant

Overview

This is a Graphical User Interface (GUI), that takes an image of the Scrabble board and returns the user a text matrix of the characters present on the board. Also, the GUI takes the images of the tiles and gives the best valid word for the given set of tiles.

Goal

With the help of image processing techniques we convert the image of the Scrabble board into a 15 X 15 text matrix.

Requirements

This project is based on the language Python 3.7.

The libraries used in this project are:

• os : Provides functions for creating and removing a directory (folder), fetching its contents, changing and identifying the current directory, etc
• openCV : Open-source library for computer vision, machine learning, and image processing
• math : Most popular mathematical functions are defined here, including trigonometric functions, representation functions, logarithmic functions, angle conversion functions, etc.
• numpy : It is a fundamental package for scientific computing in Python
• matplotlib : It is a cross-platform, data visualization and graphical plotting library for Python
• pytesseract : Optical character recognition (OCR) tool for python, which helps to recognize and “read” the text embedded in images

How To Run The Project?

Run these lines of code in your terminal to install the required libraries and code.

$ git clone https://github.com/Digital-Image-Processing-IIITH/dip-project-lena.git 
$ cd dip-project-lena/src
$ apt-get install python3-OpenCV
$ sudo apt install tesseract-ocr
$ pip install -r requirements.txt

To run just the notebook and not the GUI, run the following command when in the 'src' folder:

$ jupyter notebook main.ipynb

To run the GUI, you will need two separate terminal sessions running parallely. In one terminal, run the following command when in the 'src/webapp' folder to run the frontend:

$ cd frontend
$ npm install
$ npm start

In the other terminal, run the following command when in the 'src/webapp' folder to run the backend:

$ cd backend
$ npm install
$ npm start

Running The Notebook

Run all the cells of the Jupyter notebook present in the same order. For intermediate steps you will be able to see the outputs too. The necessary details of the functions are given in the markdown of the notebook.

Using The GUI

There are 2 parts to the GUI:

1. The first part takes an image of a Scrabble board as input and returns the extracted board in matrix form. To do this, simply click on the 1st/upper 'Browse...' button, select the board image file and click 'Open' and finally click the 'Generate Board' button. Note that this segment takes upto a whole minute to run so please be patient for the output.
2. The second part takes an image of the user's 7 tiles and returns the best word suggestion with the corresponding score. To do this, simply click on the 2nd/lower 'Browse...' button, select the tiles image file and click 'Open' and finally click the 'Generate Best Word' button.

Note: All image file formats are accepted for the above.

Results

• IN GUI

GUI Home Page

GUI Board Output

GUI Tiles/Best Word Output

• IN JUPYTER NOTEBOOK

Board 1 - Input to board segment

Output from board segment for above input

Input and Output of Tiles Segment

Please make sure you follow the project guidelines carefully.