Adaptive Real-Time Gesture Recognition in a Dynamic Scenario for Human-Robot Collaborative Applications

One of the enabling technologies of Industry 4.0 is represented by collaborative robotics. The basis of its success lies in the increase in safety, productivity and flexibility but above all in the improvement of the employees working conditions, as well as their wellbeing. Human Robot Collaboration (HRC) applications are increasing due to the ease of communication between robots and humans. To this aim, this paper presents a vision based hand gesture recognition to help humans manage the workstation operations. An application with three levels of access has been proposed to guarantee an appropriate level of user friendliness. The system provides both hardware and software support for real-time gesture recognition. The developed system was tested using a standard webcam and applied to a case study to control the tasks performed by UR10e. Thanks to its fast and on point recognition, it can be used as a real-time gesture image recognition system.