This ROS package utilizes Dijkstra's algorithm and fiducials to pilot a robot through a maze.

The project works like so:

• First, a robot moves through a maze via user-controlled teleop
• This takes a SLAM of the maze which is then saved as an image
• Dijkstra's algorithm is then used to analyze the SLAM image and output the shortest path through the maze
• Fiducials are then created in the maze based on the result of Dijkstra's which instruct what turns to make
• A second robot is placed in the maze and moves through the maze on it's own, based on the fiducials

• $ roslaunch FiducialPathFollowing maze_world_origin.launch
• $ roslaunch turtlebot3_slam turtlebot3_slam.launch slam_methods:=gmapping
• $ teleop
• $ rosrun map_server map_saver -f maze_world_origin
• $ # stop all nodes
• $ python shortestpath.py # this writes into coordinates, doesn't work
• $ roslaunch FiducialPathFollowing maze
• $ python write_launch.py # this writes into dynamic_spawn.launch
• $ roslaunch FiducialPathFollowing dynamic_spawn.launch
• $ rosrun FiducialPathFollowing follow_marker.py

• first submission: https://youtu.be/rgmT1Yg5hnU

• https://campusrover.github.io/FiducialPathFollowing/

• Jacqueline: Change Camera Pitch
• Harris: Detection of Edges