Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
audio-tests
camera_tests
gcf_ocr
output
output_old
wifi_setup
README.md
capturing_scene_audio.py
get-pip.py
intro_audio.py
intro_audio_playback.py
scheduler.py
text_label_detection.py
text_label_detection_repeat.py
text_label_detection_repeat_translate.py
translate_test.py
translation.py
vocal_focals_capturing_scene.mp3
vocal_focals_intro.mp3

README.md

Vocal Focals

A wearable text recognition device for the visually impaired.

Made using the Raspberry Pi and Google Cloud Vision API.

vocalfocals.com


This tutorial will walk through setting up a Raspberry Pi Zero W with the Google Cloud Vision API to detect text in an environment and then relay that text to the wearer over audio.

Ingredients you will need:

This tutorial will assume that you are working on a mac or linux computer.

If you have already setup your Raspberry Pi and installed git, jump to part 3 to set up VocalFocals.

1) Raspberry Pi initial setup

With a factory Pi Zero W, you can follow this tutorial to first setup the Raspberry Pi.

If you are unable to connect to wifi following this tutorial, you can follow this quick tutorial to ssh into your Raspberry Pi using USB-OTG.

Note that the factory credentials for the Raspberry Pi are:

Username: pi
Password: raspberry

2) Raspberry Pi housekeeping

You will need to do some housekeeping on the Raspberry Pi.

sudo raspi-config

In Interfacing Options, enable the camera, enable SSH, and enable VNC server if they are not already enabled.

Install VNC viewer on your computer to remotely view the Raspberry Pi's screen.

3) Install git on your Raspberry Pi

Create a GitHub account.

$ sudo apt-get install git

Configure your name:

git config --global user.name "Your Name"

Configure your email:

git config --global user.email email@example.com

4) Google Cloud Platform setup

To detect text in the image captured by the Raspberry Pi, we will use the Google Cloud Vision (GCV) API on the Google Cloud Platform. While this is a paid product, Google offers 1000 API calls for free each month, enough to experience the concept in action.

Create a Google Cloud Platform account.

...

5) Prepare the Raspberry Pi for GCV

In the /home/pi/ directory, run:

git clone https://github.com/ajzeller/vocal_focals.git

You will need to upgrade Pip, the package manager for python installations:

sudo pip install --upgrade pip
sudo apt-get install libjpeg8-dev

Next, you will install the Google API Python Client:

sudo pip install --upgrade google-api-python-client

Next, install Python Imaging Library:

sudo pip install --upgrade Pillow

Install Python PiCamera:

sudo apt-get install python-picamera

Install mpg321, a command line .mp3 player:

sudo apt-get -y install mpg321

[API key JSON stuff here]