# UAV Ground School computer vision

> Open source solutions for UAVs@Berkeley Software Ground School: computer vision for drones, object detection, and flight controller protocols, with measured benchmarks in Python and C++.

UAV Ground School is the public showcase for competition grade coursework pipelines. Week 2 covers HSV color segmentation with object centers, SUAS shape classification, and drone video mosaicking with loop closure and bundle adjustment. Week 3 covers blob detection, classical game piece detection, and YOLOv8. Week 4 covers a Cube Orange+ airframe design and from scratch protocol codecs verified on ArduPilot.

- [Week 3 object detection](https://uav-ground-school.pages.dev/week3)
- [Week 4 flight controllers](https://uav-ground-school.pages.dev/week4)
- [About](https://uav-ground-school.pages.dev/about)
- [Contact](https://uav-ground-school.pages.dev/contact)
- [Privacy](https://uav-ground-school.pages.dev/privacy)
- [Agent index (llms.txt)](https://uav-ground-school.pages.dev/llms.txt)
- [Sitemap](https://uav-ground-school.pages.dev/sitemap.xml)
- [Source repository](https://github.com/KarthikSubramanian07/UAV-Ground-School)

Built 2026-11-10. Request this URL with `Accept: text/markdown` for this representation, or `Accept: text/html` for the interactive demo page.