python-object-detection-with-ultralytics-ai-models-and-opencv

Table of Contents

• Summary
• References
• How to use

Summary

Simple Object detector app utilizing Ultralytics AI models. The detector primarily utilizes YOLOv8 and YOLO11 models.

NOTE: The initial object detector app including RTDETR models is available in the experiment folder.

References

• Inference on video frames for Ultralytics YOLO

Preconditions

• Setup Python trough an OS specific package manager (brew on Mac OS and choco on Windows) or download release from from the official website

How to use

NOTE: Tested on Windows 11 OS and Unix systems

Install pip packages

Run following commands:

# Windows 11 OS
pip.exe install -r requirements.txt

# Unix systems (NOTE: consider --break-system-packages only in case you encounter 'error: externally-managed-environment')
pip3 install -r requirements.txt --break-system-packages

Run object detector application

Run following commands on Windows 11 OS:

python.exe object_detector_app.py --help

# Object detection with pretrained nano YOLO V8 model
python.exe object_detector_app.py --model nano_yolov8

# Object detection with pretrained nano YOLO 11 model
python.exe object_detector_app.py --model nano_yolov11

Run following commands on Unix systems:

python3 object_detector_app.py --help

# Object detection with pretrained nano YOLO V8 model
python3 object_detector_app.py --model nano_yolov8

# Object detection with pretrained nano YOLO 11 model
python3 object_detector_app.py --model nano_yolov11