An applicattion on learning of five-axis robot arm with computerized visual methods

This thesis study has been developed on the ability to obtain information from large data sources such as images using a microprocessor and to process the information with a Microcontroller that will enable it to be suitable for the robot arm, which is the peripheral unit. It consists of determining the integrity or the relevant surface of the object, finding the center of gravity of the recognized area and moving it with a robot arm. The aim is to learn and process information by extracting information from image data. It is the recognition, detection and transfer of geometric objects belonging to different color spectrum with the designed and realized image processing. During the decision process, the environment variables that the image may be affected were taken under control and these variables were emphasized. In the study, image processing was focused and the basic design of the 5-axis robot arm was made. Due to the location of the upper surfaces detected in the experimental environment, the 5-axis robot arm end function is intended to approach objects at right angles, and kinematic calculations have been performed within this limitation. Since the power of the servo motors will show infinite variability during the movement of the robot arm, our experimental environment has been able to successfully supply power to the holding stage, after the capture phase the power requirement has been detected and the internal position of the gripper motor for the power to be drawn due to the motors running simultaneously during the transportation of the object, precisely. The study successfully recognized the colors and reached the specified object without any problems. The object was able to be picked up and transported in positions where power was sufficient.