Vision-based hand segmentation techniques for human-robot interaction for real-time applications

Abstract

One of the most important tasks in hand recognition applications for human-robot interaction is hand segmen-tation. This work presents a method that uses a Microsoft Kinect camera, for hand localization and segmenta-tion. One important aspect for robot control besides hand localization is hand orientation, which is used in this work to control robot heading direction (left or right) and linear velocity. The system first calculates hand po-sition, and then a kalman filter is used to estimate displacement and linear velocity in a smoother way. Ex-perimental results show that the system is easy to use, and can be applied on several different human-computer interface applications.