MEDIA

THE PHYSICAL MAZE

The cardboard maze we created to take maps of. We took two maps, one each from the robot starting at opposite ends of the maze.

MAP AND BAG

After using Hector Slam to map the maze, we ran the bag file taken when mapping, on top of the map created and were happy to see the lidar scans matched well with the map.

CORNER FINDING

Using OpenCv's Harris corner detection, we were able to find important points on the map. It successfully detected the walls to be important.

MAZE

This is the cropped map image of the second maze. This is where the matrix of points is taken from to determine where the important points are.

ROTATED MAZE

This is the important points (detected from our PointMatching code) from the above maze rotated using our translation code.

Media: Our Team