Developing the image processing based system which provides a industrial robot to work safely and cooperatively

Thanks to the interoperability (collaboration) between human and robot; flexibility, efficiency, capacity and quality in production can be increased. In this form of work (human-robot interaction), sharing the same environment between human and robot creates the possibility of collision, thus ensuring the safety in such work environments is the most important problem. In traditional robotic systems, security is provided by various equipments such as cage systems, laser scanners and light barriers. However, these systems restrict the working environment and eliminate the flexibility factor that is proposed as an advantage in the interoperability of human and robot. In this study, an image processing-based approach which is tried to be formed with the camera and hardware placed on the robot to eliminate cage and other preventive systems is presented. The camera and equipment used in the system were chosen to be of lowcost standard equipment. Thus, it has been tried to create a cost-effective and easy-to-implement solution for operating standard industrial robots in a form suitable for interaction. In the article, firstly, the method developed on multiple camera positioning and image acquisition is explained. Then, it is explained how a feature based panoramic image is obtained from the multiple images, how to make human or object detection on this image and how to make location and distance extraction via this detection. The algorithm that manages the operation of the human and the robot in the same environment using the extracted parameters and defined security rules is summarized. Finally, the advantages of the developed system compared to laser and lattice systems are discussed and predictions for future studies are given.