Yin Yang Ranch: Building a Distributed Computer Vision Pipeline using Python, O...

on YouTube ↗31m

" summary

A working hub-and-spoke pipeline runs Python on 20 Raspberry Pi computers with PiCameras that do local motion detection and send only key frames to a hub over ImageZMQ/PyZMQ and OpenCV. The hub indexes images, extracts water meter digits and keeps seasonal critter counts; a single hub handles about eight to twelve concurrent Pis at ~10–12 fps. Nodes capture up to a million frames a day but typically transmit only a few hundred, and the system reports successful longterm operation.

AI-written summary · about

" chapters (22)

1. 0:00Intro
2. 0:28Yin Yang Ranch: A Distributed Computer Vision System
3. 1:04Building a Small Permaculture Farm in Suburbia
4. 2:29Jeff Bass Bio Bullets (College, Career, Retirement)
5. 3:18How Computer Vision helps manage my small permaculture farm
6. 5:20What is a Distributed Computer Vision Pipeline?
7. 6:21My Tools for Computer Vision in Python
8. 7:35Image Computing Toolset: OpenCV
9. 8:10imagenode image ZMQ imagehub Pseudocode for imagenode
10. 12:01DCVP Example: Reading my Water Meter
11. 19:00Imagenode Code Snippet (Runs on a RPI)
12. 20:15There are many imagenode CV settings... Parameter Tuning to optimize detection
13. 20:46For each image captured, imagenode: • Applies Transformations
14. 21:35Image Transport: ZMQ
15. 23:41Imagehub Code Snippet (Runs on a Mac)
16. 25:32DCVP end to end example: the Barn Coyote Cam
17. 26:54Communicating to End Users (me :-)
18. 27:19Yin Yang Ranch DCVP: Lessons Learned So Far
19. 28:01Hardware: Example Light Types
20. 28:17Hardware Enclosures: Use Existing Light Fixtures!
21. 28:31Learnings: Some Hardware Tips and Tricks
22. 29:35Yin Yang Ranch DCVP projects are Open Source & on GitHub

" related talks

• 21mLeveraging Python for Real-Time Image processingElazar Neeman · PyCon Israel 2024
• 30mBuilding versatile operating setups for real world use and testing with Python and the Raspberry PiJens Nie · PyCon DE & PyData 2025
• 36m"Lights, camera, AI!" - Automated sports videographyZack Akil · PyData London 2017
• 16mPython in Smart FarmLex Bryan · Pycon PH 2017
• 27mHow a major museum runs on PythonGreg Turner, Ali Haberfield, Simon Loffler · PyConline AU 2020
• 33mSmart and Simple Data Driven Hydroponic Sensing Applications Using Python, Flask & Raspberry PiBethany Poulin · PyOhio 2018
• 28m"Build Your Own IoT! Fun With Python and Raspberry Pi."A.J. Fisher · PyCon AU 2024
• 41mApplying serverless architecture pattern to distributed data processingDenys Makogon · PyCon APAC 2018