COMPVIS-NERFHACKS

Autonomous Turret Track: Nerf-Bullseye-Seg Overview Autonomy pushes engineering beyond direct control and into the realm of independent decision-making. This project transforms a standard Nerf platform into a self-directed autonomous system.

By integrating Computer Vision (YOLOv8 + SAM 2.1) with precise hardware control, the turret can:

Sense its environment Interpret visual data Independently track and neutralize targets —all without human intervention.

Objectives Computer Vision Integration Real-time target detection and segmentation with a live camera feed displayed via an LED or monitor interface.

Fully Autonomous Logic A static or mobile system capable of identifying, tracking, and firing at a target autonomously.

Engineering Presentation A documented journey of system design, functionality, and iterative problem-solving.

Constraints and Compliance Dual-Axis Aiming Precise movement across at least two axes (Pitch and Yaw).

Autonomous Firing Launch sequence triggered independently based on CV “lock-on.”

Non-Invasive Modification Strictly follows the no-modifications rule — the Nerf gun is fired using external actuators only, with no changes to the original firing mechanism or pellets.

Durability Built to withstand repeated testing and high-vibration firing sequences.

Tech Stack Detection: YOLOv8 (real-time object detection) Segmentation: SAM 2.1 (Segment Anything Model) Safety Logic: Integrated human-avoidance protocols (skips processing if cls == 0) Language: Python 3.11+ Repository Structure nerf-bullseye-seg/ ├── nerf_detector.py # Core CV detection and tracking logic ├── Instructions.md # Hardware setup and calibration guide ├── requirements.txt # Python dependencies └── .gitignore # Excludes heavy model weights (.pt files) Created By BU Women in Data Science (WiDS)

Connect With Us

Instagram: @bu_wids Email: [email protected]

license: mit