Development of a route planning system for mobile robots using visual data

With the current development of computer system technologies and its overlap with many fields of science and technology, and as a result of current quantum leap of computer vision applications, the development space became open for variety of applications. In this thesis computer techniques are used for an application in control systems and robotics, which is related with observing a mobile robot in a given environment, recognizing the robot and discovering the obstacles in the environment using several methods in the fields of image processing, object recognition and object tracking. An improved algorithm is developed to generate the best route using optimal search algorithms, it is called Omni-Directional Adaptive A*. An application program on a computer device continuously reads the visual data from the camera, processes it, and generates the best route. Sending commands via a wireless network, developed system guides the robot to follow the best route avoiding the obstacles and reach the target using shortest way, which helps to reduce power and time consumption. The developed system could be useful in a variety of applications including industrial automated logistics, search and rescue robotics, transportation of materials inside huge factories and warehouses.