Camera Pi

This repository allows for privacy masks to be enabled and disabled. It is intended to be connected to a security system or some system that can provide an "armed" state and an "alarm" state.

security-system-check.py Flowchart

Environment Setup and Pi Setup

With a Raspberry Pi and a base OS perform the following steps to deploy the code. You should be able to use any Raspberry Pi. In the next steps you will clone the git repository, setup your Python virtual environment, download the required packages from PyPi, and then install the service. This assumes you are currently in the user's home directory and with a subfolder of security-camera-privacy-mask

1. git clone git@bitbucket.org:mount7freiburg/security-camera-privacy-mask.git
2. cd security-camera-privacy-mask
3. python -m venv .venv
4. .venv/bin/pip install -r requirements.txt
5. cp .env.example .env
6. cp cameras.json.example cameras.json
7. Make appropriate changes to .env file
8. Make appropriate changes to cameras.json file
9. Connect wires to GPIO 16 and GPIO 20 for the Raspberry Pi
10. Configure crontab to run deploy.sh (optional)
    1. This will perform a git pull and restart the service

Service Setup

1. cp security-camera-privacy-mask.service /etc/systemd/system/camerapi.service
2. sudo systemctl daemon-reload
3. sudo systemctl enable camerapi.service

Understanding the env vars and cameras.json file

I am using smtplib and Gmail for sending emails. You can use any service you want. If you use a different service you will need to probably configure a different client or SMTP service.

Important

In Dahua cameras the password has a maximum of 32 characters even though the form lets you input more. Keep the .env at 32 characters if your password is longer and it should work

The .env file also contains LOGFILE_PATH should be the full path to your log file. LOG_DEBUG can be any integer that is typically used for Python logging levels (0, 10, 20, 30... etc). CAMERA_USERNAME and CAMERA_PASSWORD needs to be a user that is allowed to make configuration changes. The camera API calls are programmed to call Dahua cameras. This should also work with Dahua whitelabeled cameras.

The cameras.json file contains channel, type, and ip. Only ip and type are required. The type field can either be exterior or interior. Channel doesn't do anything but can be used as a reference for you.

Testing

For testing, at the moment, I just comment out the GPIO lines and run my tests by changing the booleans to test the cameras. The most common problem I encountered when testing was if the cameras are on a different subnet and the lengthy password preventing the API call to fire.