This example demonstrates Object Detection with oriented boxes using the off-the-shelf YoloV8s-OBB model from Ultralytics compiled and running on the accelerator. It implements an object detection pipeline for oriented bounding boxes (OBB) where:
- Objects are detected in an image.
- Bounding boxes and keypoints are generated to represent detected objects.
- The bounding boxes and keypoints are processed for further use in downstream tasks.
Before running the application, ensure that all requirements are installed.
pip install PyQt5
pip install opencv-python==4.11.0.86| Property | Details |
|---|---|
| Model | YoloV8s-OBB |
| Model Type | Object Detection (Oriented Bounding Boxes) |
| Framework | Onnx |
| Model Source | YoloV8s-OBB |
| Pre-compiled DFP | Download here |
| Output | Object bounding box + keypoints |
| OS | Linux |
This project uses third-party software, models, and libraries. Below are the details of the licenses for these dependencies:
- Original Idea: Traffic Analysis Example from Roboflow Supervision
- License: MIT License 🔗
- Models: YoloV8s-OBB Model exported from the Ultralytics GitHub Repository
- Default Video: Default Drone video 🔗
- License: Pexels License 🔗
This example implements an Object Detection with oriented boxes, utilizing the off-the-shelf YoloV8s-OBB model. It showcases how to detect, track, and analyis traffic.
HOSEN ARAFAT
Software Engineer, China
GitHub: https://github.com/arafathosense
Researcher: Artificial Intelligence, Image Computing, Image Processing, Machine Learning, Deep Learning, Computer Vision