garage-monitoring

The repository contains the source code for Smart garage control with Firebase AutoML and RaspberryPi.

Here we will explore a creative way to monitor our garage doors - using a Raspberry Pi, Firebase, and a web camera, with some magic using machine learning to send us reminders when we forget to close our garage door.

Step 1: data collection

The take_picture.sh script that will take a picture and save it.

#!/bin/bash

DATE=$(date +"%Y-%m-%d_%H%M%S")

fswebcam -S 30 -r 480x240 --no-banner /home/pi/Pictures/data-collection/$DATE.jpg

Then we set up a cron task using crontab -e to run this script every 5 mintue.

*/5 * * * * bash ~/Documents/github/garage-monitoring/data-collection/take_picture.sh 2>&1

We can also upload all pictures to a Google Cloud Storage bucket every 30 minute.

*/30 * * * * gsutil -m cp -r /home/pi/Pictures/data-collection/ gs://garage-door-training-data

Step 2: Train the model with Firebase AutoML

Refer to the article for details.

Step 3: Using the model

First we setup another cron task to run get_live_view.sh, which take a picture every minute.

* * * * * bash ~/Documents/github/garage-monitoring/get_live_view.sh 2>&1

Then we run the app.py script which uses this picture to determine if our garage doors are closed, and send us an SMS reminder if they are not.

Appendix

Installing tensorflow 2.0 on Raspberry pi

If pip3 install tensorflow doesn't correctly install tensorflow, try to following:

$ sudo apt-get install -y libhdf5-dev libc-ares-dev libeigen3-dev
$ sudo pip3 install keras_applications==1.0.8 --no-deps
$ sudo pip3 install keras_preprocessing==1.1.0 --no-deps
$ sudo pip3 install h5py==2.9.0
$ sudo apt-get install -y openmpi-bin libopenmpi-dev
$ sudo apt-get install -y libatlas-base-dev
$ pip3 install -U --user six wheel mock
$ sudo pip3 uninstall tensorflow
$ sudo -H pip3 install tensorflow-2.0.0-cp37-cp37m-linux_armv7l.whl

【Required】 Restart the terminal.

For detail see this repo.