An Image Processing Approach to Solve Labyrinth Discovery Robotics Problem

摘要

Maze solving using multiple algorithms is one of the important problems in the last years, maze solving problem is to determine the path of a mobile robot from its source position to a destination position through a workspace populated with obstacles, in addition to finding the shortest path among the solutions. Autonomous robotics is a field with wide-reaching applications, from bomb-sniffing robots to autonomous devices for finding humans in wreckage to home automation; many people are interested in low-power, high speed, reliable solutions. we will introduce an algorithm to solve mazes and avoid robots to go through the long process, this algorithm is based on image processing and shortest path algorithm, the algorithm will work efficiently because of the preprocessing on maze's image data rather than going through the maze cell by cell. Therefore, it focuses on the entire maze rather than the current part that an agent is in it. This approach will give the robot preplanning time before falling in mistakes like loops and being trapped in a local minimum of dead ends of labyrinth besides saving time, because the entire maze map will be known before the navigation starts.