Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/edwardjxli/wormhole

A Simple and Hackable Realtime Video Streaming Engine for Prototypes and Projects Alike!
https://github.com/edwardjxli/wormhole

camera-streaming http-live-streaming opencv python python-video-streaming realtime sockets streaming video video-streaming video-streaming-server

Last synced: 3 months ago
JSON representation

A Simple and Hackable Realtime Video Streaming Engine for Prototypes and Projects Alike!

Awesome Lists containing this project

README 

        
[HTMLElementBegin]: # 



  

  

  Wormhole Logo



[HTMLElementEnd]: # 



# Wormhole: Realtime Streaming Engine

Wormhole is a simple and hackable realtime video streaming engine for prototypes and projects alike! Stream amera feeds, video files, or machine learning output from anything like a Raspberry Pi to a fully fledged development environment!



![wormhole_screenshot_spring](https://user-images.githubusercontent.com/20020059/184041754-5ee78d89-6796-49b8-87b7-13792c301d83.jpg)



> ## Try Out A Live Demo Here: https://demo.wormhole.hydranet.dev/



## Features:

- Multithreaded Flask & SocketIO backend

- Camera and Video File Streaming!

- Fast and scalable Motion JPEG and PNG streaming

- Dynamic protocol negotiation between Wormhole servers and clients

- Dynamic video modification through callbacks and views



## What is Wormhole?

Wormhole is not meant to overtake established battle-tested libraries such as WebRTC or alike. Instead, Wholehole aims to be an extremely simple-to-use library wherever you need to share video (Pre-recorded or live from a camera) between two or more servers. The original design goal was so that Wormhole could be utilized during fast-paced development environments such as hackathons and other competitions, yet flexible and modular enough to be a one-stop-shop for your next AI video processing project, webcam streaming project, or simple video sharing project. From a Raspberry Pi camera server to annotated output from a Machine Learning pipeline, Wormhole can fit all your needs!



## Installing Wormhole

To get started with Wormhole, you can just clone the project by running `git clone [email protected]:RadioactiveHydra/Wormhole.git` to your local machine.

Once downloaded, run `pip install -r wormhole/requirements.txt` to install all the required dependencies.

> NOTE: It is recommended that you install Wormhole in a virtual env! More instructions can be found [here](https://www.freecodecamp.org/news/how-to-setup-virtual-environments-in-python/)



## Getting Started with Wormhole

> NOTE: This is still WIP! API may change and/or be removed in future releases.  



To start streaming from your camera, all you need to write is:

```py

from wormhole import Wormhole

server = Wormhole()

server.stream(0)

server.join()

```



**Thats It!** Wormhole then handles setting up the server, launching the camera, and streaming the video frames!

To stream from a video file instead, you can write the following:

```py

from wormhole import Wormhole

server = Wormhole()

server.stream("path/to/video.mp4")

server.join()

```



As easy as that! To view the video stream in another python project, you can just run:

```py

from wormhole import Wormhole

video = Wormhole().view("your.ip.address.here")

```



Again, **thats it!** Wormhole then handles setting up the client, negotiating the best streaming protocol to use, and viewing the video stream! To render the video to your screen, you can then run:

```py

from wormhole.utils import render_video

render_video(video)

```

The video stream is also available online if you go to `http://localhost:5000/wormhole/stream/default/mjpeg` in your browser.



Of course, you may want to do more than what Wormhole offers by default. For that, you can check out some examples in the `examples` folder or in the [Official Wormhole Example Server](https://github.com/EdwardJXLi/WormholeExampleServer)



## Why develop Wormhole?

After rewriting the same video streaming codebase for the 5th hackathon/project in a row ([Ref.1](https://github.com/CrowdEye/crowdeye-ai-engine/blob/bbb1591cbc121babef8de973ba7210fe358683ba/ai.py#L57)) ([Ref.2](https://github.com/MaskPass-BellCSclub/ai-server/blob/1d9acdc36677fa433a0f3db2c2d297fe023c0d70/main.py#L199)) ([Reg.3](https://github.com/Hive-HQ/Hive-HQ-backend/blob/27f88a2a74da9465f8fc1ceb89fd0344d696a8cc/main.py#L526)) (& More!), I thought it was time to sit down and write a universal streaming library for all my needs. Shopping around online did not yield any projects that fit my requirements, so I decided to build my own. Wormhole is designed so that it can not only be deployed in ~~minutes~~ **seconds**, but it is also flexible enough to do everything I need in a fully-fledged data streaming library. 



## Contributing to Wormhole

Wormhole's source code follows all **PEP8** guidelines (except E501 for long lines). Also, most function arguments are **typed** so that the API is coherent and easy-to-use. Finally, Wormhole aims to keep compatibility for **Python versions 3.8 and beyond**. To contribute to Wormhole, you can either fork the project on GitHub or submit a pull request. If you have any questions or suggestions, feel free to join the discord server here: [https://discord.gg/9QF2bPc](https://discord.gg/9QF2bPc).



[HTMLElementBegin]: # 

## Powered By Wormhole Assets



  

  

  Powered By Wormhole



[HTMLElementEnd]: #