RGB-D sensor based SLAM and human tracking with Bayesian framework for wheelchair robots

In this paper, we present an approach to visual SLAM and human tracking for a wheelchair robot equipped with a Microsoft Kinect sensor that which is a novel sensing system that captures RGB and depth (RGB-D) images simultaneously. The speeded-up robust feature (SURF) algorithm is employed to provide the robust description of feature for environments and the target person from RGB images. Based on the environmental SURF features, we present the natural landmark based simultaneous localization and mapping with the extended Kalman filter suing RGB-D data. Meanwhile, a depth clustering based human detection is proposed to extract human candidates. Accordantly, the target person tracking is achieved with an online learned RGB-D appearance model by integrating histogram orientation of gradient descriptor, color, depth, and position information from the body of the identified caregiver. Moreover, a fuzzy based controller provides dynamical human following for the wheelchair robot with a desired interval. Consequently, the experimental results demonstrated the effectiveness and feasibility in real world environments.