A shortest-path maze solving algorithm for image-based mazes.

maze, algorithm, image, solve, algorithms, easy-to-use, image-manipulation, image-processing, image-recognition, shortest-path
pip install mazesolver==2.2.2


Maze Solver

This algorithm finds the shortest path through black and white maze images. It outputs an image with the path marked in green.


Check that your python version is >=3.5 with python3 -V. Also make sure that pip is installed with python3 -m pip -V.

To install, simply run python3 -m pip install mazesolver --user on the command line.

You can run the tool using mazesolver or mazesolver.exe.

What are the rules for maze images?

  • Walls marked with black pixels and paths marked with white pixels

  • Walls around the entire maze

  • One entrance on the top row and one exit on the bottom row

You can generate a compatible maze using mazegenerator, or just make your own image in an image editing program.

If this is confusing, check out some of the example mazes in the pics directory.

How do I use it?

You can use mazesolver --help to get a list of commands.

Make sure that your image meets the specifications above.

Normal usage will look something like this: mazesolver -i path/to/input_img -o path/to/output_dir/

How does the algorithm work?

  • Start at the entrance and label it zero.

  • Move to any neighbouring cells and mark them with 1

  • Move to any of the cells neighbouring the ones marked 1 and label them 2

  • Continue doing this until all cells are marked

  • Start from the exit of the maze and move to any neighbouring cell that == the current cell's number -1. Until we reach 0 (the entrance).

We now have the shortest path from the entrance to the exit!