FROM IMAGE TO GRAPH-BASED DATA FOR PATH PLANNING

Path planning nowadays have been widely used even in our daily lives such as navigation of Google Maps or Waze. Path planning is one of methods that have been used in finding ways from one point to another. There’s varies approach to be used for path planning and one of the approaches that will be discuss in this report which is a study of graph-based data from an image of a maze-solving path project. This implementation of graph traversal, statistics and machine learning algorithms were a challenging part to be done. Input will be taken as an image and the process of converting the image to become a function by graph-based data methods have been taken as the important points to succeed this project by using the Image Processing Toolbox in MATLAB. This toolbox is used to filtering with the given pixels is encircling by the nearby pixels to be formed as the neighborhood and convert it into a graph with a set of matrices. This procedure also will then can be implemented to the robot for finding the free pathways in future project. Thus, the comparison of different methods of Dijkstra algorithm that have been used in past has been narrowed down to the subsystem that will be the guidance of this project to find the shortest pathway.