Robomaster EP: Perception, Planning, and Navigation

This repository contains a full robotics navigation stack developed for the Robomaster EP platform. The project integrates computer vision (AprilTags), graph-based path planning (Dijkstra, BFS, DFS), and autonomous trajectory execution in a structured environment.

Overview

The system is designed to navigate a robot through a maze of 26.6 x 26.6 cm storage cubesusing markers for localization.

Main files

• apriltag.py: Real-time vision pipeline that detects tag36h11 AprilTags to estimate the robot's 6D pose relative to landmarks
• final_chunk.py: A sequence-based maze solver that executes decomposed path segments while maintaining yaw alignment via visual feedback.

Videos

• AprilTag Tracking
• Maze Solving