This is an implementation repository of dimension integrator. This repository solves the problem statement "integrate robot dimensions into path planning" The algorithm uses computer vision technique to alter the map in a manner which would close all the passages which are smaller than the robot width.

To run the program in the repository you will see a Python file named constants.py, in this file you will provide the path to the binary map and where the output should be stored.

robot_width = 20 #enter the robot width in pixels
name_BM = "sample_map_1"
path_to_original_BM = str(r"path to input directory/"+name_BM+".png")
path_to_processed_BM = str(r"path to output directory/"+name_BM)

Above is a code excerpt demonstrating how to use the algorithm. The file optimization.py includes the entire algorithm which will process the 2D-Binary map and store the processed map to the given path. to use this algorithm with the path planning algorithm you can import the function dimension_integrator from optimized_pipeline.py and give it the input binary map and robot width in pixels as arguments. Example of this is given below:

from optimized_pipeline import dimension_integrator
processed_binary_map,result_visualization = dimension_integrator(map,robot_width)

Experimentation videos can be found in the folder of experimentation results.