Skip to content

Latest commit

 

History

History
171 lines (117 loc) · 5.9 KB

README.md

File metadata and controls

171 lines (117 loc) · 5.9 KB
Raw

pet-etc-monitor: An application for monitoring from cheap ESP32 CAM camera boards

This pet-etc-monitor application is written largely in Python using Flask to make more effective use of the cheap and ubiquitous ESP32 CAM Wi-Fi enabled camera boards. See http://www.ai-thinker.com/pro_view-24.html for the ESP32 CAM datasheet. These camera boards (or their clones) are available for less than $10 from some sources such as AliExpress.

When monitoring a pet with an ESP32 CAM, it was found that while using the firmware from examples for the Arudino IDE that the video streaming is unreliable and often freezes up. It was often hard to tell if an image was frozen or not.

Features of this application include:

  • Show images as they come in. The ESP32 CAM is often seconds per frame so this is not a significant drawback for simple monitoring.
  • Show multiple images, some with a user configured delay so what happened at past times can easily be compared to the present.
  • Show warnings (through colour change) when images are unexpectedly delayed.
  • Stores images on disk with metadata in a PostgreSQL database for future examination.

Monitoring plants to make sure the cat doesn't get into them

Monitoring plants to make sure the cat doesn't get into them

ESP32 CAM hardware and firmware

This application currently uses as firmware the ESP32 CAM examples for the Arudino IDE.

  • Install the ESP32 board manager according to the instructions at https://docs.espressif.com/projects/arduino-esp32/en/latest/installing.html.

  • After switching to AI Thinker ESP32-CAM under Tools-->Boards-->esp32-->AI Thinker ESP32-CAM, use the CameraWebServer example will be located under the File-->Examples-->ESP32-->Camera menu.

  • This can then be uploaded to the ESP32 Cam camera board using the usual procedure for the Arudino IDE.

IP address of the camera

A router that reserves an IP address for the ESP32 Cam can be used.

Installation

This application consists of a recording application and a viewing application. Both have a browser-based interface and through appropriate configuration these can be on different systems. This application has been tested on a Debian Linux 12 system.

Installing a Python virtual environment

On Debian 12 the packages corresponding to the following command probably need to be installed

sudo apt-get install python3 python3-venv

To install and use a virtual environment in the project directory, follow the instructions on the Flask website https://flask.palletsprojects.com/en/2.3.x/installation/.

Some commands to build (or rebuild) the default virtual environement are:

python3 -m venv ./venv source venv/bin/activate pip3 install -r requirements.txt

Configuration

The application is configured through the config_test.json file.

In the config_test.json file the address and port of both the recording application and viewing application can be configured. The rotation angle in degrees of the images of the camera can be configured too.

Creating the database and table

Create a PostgreSQL database following the instructions for your operating system and PostgreSQL. This package was developed on Debian 12 so the instructions at https://wiki.debian.org/PostgreSql work well.

In config_test.json configure POSTGRES_DBNAME, POSTGRES_USER, POSTGRES_PORT, POSTGRES_HOSTNAME, POSTGRES_PW, POSTGRES_IMAGE_TABLE. See the placeholders in the afformentioned file for descriptions.

Configuring the ESP32 Cam

In config_test.json add the IP address and/or URL of the ESP32 Cam to CAPTURE_URL, STATUS_URL, and CONTROL_URL.

Running the test servers

In the project directory with the appropriate virtual environment is activated using source venv/bin/activate, run the following commands each in their own terminal.

./test_record_daemon.py

./test_control_app.py

./test_view_app.py

The default URL for the viewing application is http://localhost:8080. The default URL for the recording application is http://localhost:8081. These can be configured through config_test.json. The database table can be deleted (DROPed!!! Careful!!!) by running the command.

./test_record_daemon.py --reset-database

When first run the test_record_daemon.py script will create the table specified in config_test.json.

Convienience scripts are also provided in test_control_app.sh, test_record_daemon.sh, and test_view_app.sh that will restart the relevant program on a signal.

Running the test servers using WSGI

In the project directory with the appropriate virtual environment is activated using source venv/bin/activate, run the following commands each in their own terminal.

./test_record_daemon.py

mod_wsgi-express start-server --port 8081 test_control_app.wsgi

mod_wsgi-express start-server --port 8080 test_view_app.wsgi

As above, the default URL for the viewing application is http://localhost:8080. The default URL for the recording application is http://localhost:8081. These can be configured through config_wsgi.json.

Future Improvements

Things I would like to improve in this application.

  • Greatly reduce the latency from camera to database to web browser.

  • I am currently modifying the ESP32 CAM firmware to use MDNS.

  • Better inspection of past images.

  • Better management of the database and disk space from the browser interface.

  • Add https and logins to help secure the application.

  • Add additional sensors such as a temperature sensor.

  • Add controls for things like an LED to light up the scene.

  • Add permissions for the recording app.