Collaborative robotics offers new ways for human-robot collaboration, for which it gained a large amount of popularity in recent decades. The use of robots is spilling from industrial environments to homes, schools, etc. Robots are designed to automate processes and offer help in many ways. Despite their indisputed capabilities, many people still distrust the use of robots. Until now the main users were programmers and engineers, mainly due to complex and non-intuitive controls. In this thesis we present a new approach to controlling robots in a natural and interesting way, hoping to popularize their use. We write a program for controlling an industrial robot via hand-tracking by using images captured with a camera. We test the approach by using a collaborative robot Fanuc CR-7iA/L and a basic built-in camera. To write the software we use the Python programming language, relying on MediaPipe and OpenCV libraries for hand recognition. Even a minimalistic approach offers sufficient control to create simple drawings with the robot.
|
