Repository navigation
PixPick v0.2.0 – Multi-Region Selection
🌟 Summary
PixPick v0.2.0 introduces one of the most requested features: multi-region selection. You can now draw and manage multiple bounding boxes and polygons within a single session, making interactive region selection significantly more efficient for computer vision workflows.
📊 Key Changes
- ✨ Added support for selecting multiple bounding boxes
- ✨ Added support for selecting multiple polygons
- ⚡ Improved the interactive selection workflow for multi-region use cases
- 🛠️ General bug fixes and internal improvements
🎯 Purpose and Impact
Many computer vision applications require working with multiple regions in the same image. Previously, users had to restart the selection process for every new region.
With v0.2.0, PixPick enables seamless multi-region selection, reducing repetitive interactions and improving productivity for tasks such as:
- Object Detection (YOLO)
- YOLOE visual prompt
- SAM / SAM2 prompts
- Region-based analytics
- OpenCV applications
- Supervision PolygonZone
This release makes PixPick a more efficient utility for developers building interactive computer vision pipelines.
What's Changed
New Features
- Support for selecting multiple bounding boxes in a single interaction.
- Support for selecting multiple polygons in a single interaction.
- Improved APIs for handling collections of selected regions.
Improvements
- Enhanced selection experience for multi-region workflows.
- Internal code cleanup and stability improvements.
- More detailed docs about drawing control keys
- Separate color for box, polygon and line
Contributors
Thank you to everyone who contributed ideas, reported issues, tested new features, and helped improve PixPick your efforts helped make v0.2.0 possible.
Thank you for using PixPick! We appreciate your feedback and contributions. If you find the project useful, consider giving it a ⭐ on GitHub.