Beating My Girlfriend At Scrabble With Raspberry Pi!

Intro

So you want some help beating your friends or family at Scrabble? Well this may be the answer.

Using just a bit of Python code, running on a Raspberry Pi with a HQ Camera Module, you can read your scrabble tiles, and have a display show you the top highest scoring words that you could play.

Video Tutorial

Here's the video I made for my Devscover YouTube Channel:

Gathering Components

First you are going to want to gather all your components. Note the High Quality camera isn't essential, you can use the older camera modules for Raspberry Pi. The advantage this new one gives you, is that it has a really good depth of field, so you can have your tiles in focus while the background is blurry.

Connect It Up

1. Connect the display to the Raspberry Pi's DSI Screen Port, and required jumper cables as per the included instructions
2. Connect the HQ camera to the Raspberry Pi's Camera port (this can look the same as the DSI one)
3. [Optionally] Attach a tripod to the camera for easier mounting
4. Get out your scrabble board and some tiles

Install Dependencies

Install Pytesseract

This assumes you have Python 3 installed.

Installation is complicated, because the default way of installing should be

sudo apt-get install tesseract-ocr

however, that installs (at time of writing) version 4 of tesseract, which does not work with a whitelist.

Many attempts were made to use --oem 0 with tesseract 4 to enable whitelisting, but then more errors sprang up about needing dictionaries.

Therefore the solution was to compile the library locally from the source using these commands:

• git clone --depth 1  https://github.com/tesseract-ocr/tesseract.git
• cd tesseract-ocr/
• ./autogen.sh
• ./configure
• make
• sudo make install
• sudo ldconfig
• tesseract -v

Then you must download the english trained data from: https://github.com/tesseract-ocr/tessdata/blob/master/eng.traineddata

Next, set an environment variable to the location of that english trained data.

You can use export TESSDATA_PREFIX=/home/pi/tesseract/tessdata (replacing the location with the location of the folder you downloaded the traineddata into) in your CLI or add this to your ~/.bashrc to ensure it is set on startup

Install Python packages

You also need to use pip3 to install required packages. Use this command

pip3 install opencv-python==3.4.6.27 pip3 install pytesseract pip3 install time pip3 install PySimpleGUI pip3 install Image

Get the code

You can find the code here: https://github.com/wazcov/devscover-youtube/tree/master/scrabble

Git clone to download the repository to your Raspberry Pi

Modify the code

Next, you might want to alter a few bits to get it working for you...

Specifically this line  that alters the point at which it decides what is black and what is white

blackwhite = img.point(lambda x: 0 if x < 166 else 255, '1')

the 166 worked very well for the light source I had. However, as the day progressed the light changed and it made pytesseract unable to read the tiles. This number is therefore very fickle - you may need to do a lot of tweaking to get it to read the image.

Line up the image

To line up the image, I recommend using raspistill. This will preview for 15 seconds and save a picture to testImage.jpg

Run It!

Simply runpython3 scrabbleWordFinder.py

How did it go?

Let me know here or as a comment on my YouTube Channel, and I'll try and get back to you with tweaks and advice! Also, do subscribe on YouTube for plenty more cool tutorials like this!

You can also find me on my website Devscover.com