YOLOv11 Zenoh

ZENOH YOLO MODEL - Zenoh-based Object Detection

A deep learning-based YOLOv11 object detection project utilizing Zenoh protocol for efficient communication and data streaming. This implementation supports real-time detection from multiple input sources, including cameras, video files, and images. Developed and tested in a WSL (Windows Subsystem for Linux) environment. Written in Python and C++ for a university course.

Features

• ✅ Live Camera Detection: Process camera frames in real time via Zenoh.
• ✅ Video & Image Detection: Detect objects in video files or images.
• ✅ Multi-class Support: Recognizes 22 object classes, including PASCAL VOC 2012 dataset objects and custom classes like student and employee ID cards.
• ✅ Zenoh Messaging System: Sends detection results via Zenoh topics, making it suitable for distributed and edge AI applications.

Clone the repository

git clone https://github.com/SlayM4yd4y/zenoh_yolo_model.git
cd zenoh_yolo_model
pip install -r requirements.txt 

You will also need:

YOLOv11

git clone https://github.com/ultralytics/ultralytics.git 
cd ultralytics
pip install -r requirements.txt  

Zenoh

git clone https://github.com/eclipse-zenoh/zenoh.git
cd zenoh-python

Run

For more details visit the /wiki of this repository.

Diagram

graph LR;

train([ train.py]):::red --> pubsub[ zenoh_pubsub.py]:::light
detector([ detector.py]):::red --> pubsub
detector --> conv([ http_to_zenoh.py]):::red --> cp([ camera_pub.py]):::red --> pubsub
cg([ card_gen.cpp]):::light
ca([ card_augmenter.cpp]):::light
ic([ identifier_cleanup]):::light
gpp([ get_package_path.cpp]):::light

classDef light fill:#34aec5,stroke:#152742,stroke-width:2px,color:#152742  
classDef dark fill:#152742,stroke:#34aec5,stroke-width:2px,color:#34aec5
classDef white fill:#ffffff,stroke:#152742,stroke-width:2px,color:#152742
classDef red fill:#ef4638,stroke:#152742,stroke-width:2px,color:#fff

Loading