![]() | Only 14 pages are availabe for public view |
Abstract The goal is to apply vision based robot control by exploring the information provided by vision sensor in feedback control loop. This visual based controller is used to control the motion of designed four wheels mobile robot in unstructured indoor environment. Tasks performed by the robot required the robot to move from initial position toward a desired position while keeping the object in the field of view of the camera with the capability to avoid collision with front obstacles encountered during its motion. A mobile robot is designed with suitable dimensions and load capabilities to be able to carry a laptop and a holder for a camera. The designed robot is a four wheeled having one motor for each wheel and equipped with two infrared proximity reflection switch sensors to be able to avoid front obstacles during its mission and equipped also with Microsoft Kinect camera to track the target and give the feedback signal to the robot controller. In order to achieve this goal, LabVIEW software is used to implement a vision-based controller which consists of the integration of three sub-systems, vision subsystem, PID controller subsystem, and avoidance subsystem. Vision and avoidance subsystem send their feedback signal to the controller subsystem to correct the errors in the robot motion. This vision based controller allows the robot to perform visual navigation and follow target object. The vision- based controller system with obstacle avoidance has been investigated experimentally using the designed four wheeled mobile robot. |