RGB-D sensor-based visual target detection and tracking for an intelligent wheelchair robot in indoors environments

Abstract

The ability to follow or move alongside a person is a necessary skill for an autonomous mobile agent that works with human users. To accomplish such tasks, we develop a new scheme of visual-target detection and tracking for a wheelchair robot equipped with Microsoft Kinect that captures RGB images along with per-pixel depth information (RGB-D camera). The speeded-up semi-supervised online boosting algorithm is employed to provide the robust description of feature for environments and the target person from RGB images. Based on the environmental Haar-like features, we utilize an extended Kalman filter (EKF) based localization to estimate robot pose. Then obstacle avoidance navigation approach based on ŋ 3 spline trajectory planning and optimization are proposed for the wheelchair robot. Finally, the experimental results are provided to demonstrate the effectiveness and feasibility in real world environments.