Theoretical 3-D Monitoring System for Human-Robot Collaboration

Abstract

Three-dimensional monitoring systems, based on sensor-fusion of Microsoft Kinect 3-D cameras, can create a safe workspace for human-robot collaboration. Cameras fields of view overlap to create a higher frame rate within the workspace which allows for more accurate and quicker responses to humans in an industrial environment. An experiment was also conducted to determine the computational cost of a Kinect monitoring system under various loads, focusing mainly on CPU and GPU utilization. The experimental results showed that 3-D image processing can be processed in parallel benefitting from the higher core count found on modern processor.