https://github.com/parlaynu/pi-multicam

RaspberryPi camera toolkit and remote viewer for multiple cameras.
https://github.com/parlaynu/pi-multicam

opencv picamera2 qt-qml raspberry-pi raspberrypi-camera zenoh zeromq

Last synced: about 1 month ago
JSON representation

RaspberryPi camera toolkit and remote viewer for multiple cameras.

• Host: GitHub
• URL: https://github.com/parlaynu/pi-multicam
• Owner: parlaynu
• License: mit
• Created: 2025-04-20T02:29:28.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2025-06-08T04:05:24.000Z (about 1 year ago)
• Last Synced: 2025-06-08T05:39:06.806Z (about 1 year ago)
• Topics: opencv, picamera2, qt-qml, raspberry-pi, raspberrypi-camera, zenoh, zeromq
• Language: Python
• Homepage:
• Size: 4.34 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Multi-Camera Capture and Viewer

This repository contains a toolkit to build dataflow pipelines for recording and processing images 

from RaspberryPi cameras, sending them over a network, and a video viewer tool that can display 

image streams from multiple sources at once.

## Demo

A simple setup would consist of the following devices:

    

        

        

        

    

    

        pi-cam01

        pi-cam02

        laptop

    

To run the system, first start the cameras running on each of `pi-cam01` and `pi-cam02`. You will need

to install the software and configure the systems (documentation coming), and once that

is done, simply run this command on each raspberry-pi with the cameras:

    $ ./camera/ck-run cameras/configs/zenoh/zenoh.yaml

On the laptop, once you have built the `multiview` application (documentation coming), you can view the images by

running this command:

    $ ./multiview/build/multiview home/pi-cam01/image home/pi-cam02/image

The applications use a technology called [zenoh](https://zenoh.io/) to communicate. You just need to

make sure the computers are all running on the same network, and if you provide the correct logical

locations of the image servers, then `multiview` will find them itself - no need to specify IP addresses and

ports.

This will result in displaying two image streams side by side as shown in the screen capture below:

![Image Viewer](docs/multiview-capture.png "Image Viewer")

The `mulitview` window is shown here as it appears on a Mac. The application is cross platform and can be built

and run on a RaspberryPi and Windows as well.

## Tools

### Camera Toolkit

* [Setup and Usage Instructions](docs/build-camera.md)

* [The Details](docs/details-camera.md)

### Multiview

* [Build and Usage Instructions](docs/build-multiview.md)