RGB-D Sensor-based Visual Target Detection and Tracking for an Intelligent Wheelchair Robot in Indoors Environments

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