Special Gesture Recognition And Expansion Of Cloud Applications For Home Service Robots

Besides voice based robot-human interaction, home service robots also require "body language" to assist its communication with humans. Therefore, gesture recognition is very necessary for a robot. The human posture is ever-changing, is difficult to obtain a stable and general of gesture recognition algorithm. Aiming at the common gusture issues of wave and fall in family life, wave and lying person detection approach were raised respectively, furthermore,for the extension of the methods and heterogeneous robot frequently face similar tasks, a framework of cloud robot is proposed for sharing skills. The main contents and contributions of this thesis are as follows:(1) Locating a hand for wave detection is a challenging problem. A hand detection approach based on face recognition is proposed. The method first based on the face recognition algorithm to get the position of the face, through image processing using the Flood Fill Algorithm labeled the connected region, then constructing an undirected graph to get the finger’s location. Finally, according to the accumulated moving distance of the finger to judge whether the occurrence of waving. Experiments show that our approach meets the wave detection requirements for service robots and has been demonstrated to be practical and reliable, even when there have other people’s interference..(2) Human detection is a basic functionality for home service robots. For complex family environments where lying person is partially occluded or in cluster, we propose a lying person detection approach integrates 3D point cloud segmentation and local feature matching. Our approach segments the point cloud of each object into several pieces, matches local features of each object piece, and classifies them to detect lying person. Experiments show that our approach meets the human detection requirements for service robots and has been demonstrated to be practical and reliable, even when parts of human body are occluded.(3) To address the complexity in gesture recognition and the repetitive mission problem of heterogeneous robots, a skill sharing method based on cloud robots is developed. Our system differs from other systems, in that, it allows robots to share information and learn new skills using the network in several aspects:(1) Our system uses SNL to communicate between robots, which is based on the observation that most service robots have the ability to interpret NL sentences. This leads to better flexibility, adaptability to different robots, robustness, and the ability to reuse knowledge once acquired. (2) Humans and web accessible knowledge bases can make use of the system as they were robots. (3) Service robots can interact with each other through the framework with minimal changes in their systems. We also provide methods enabling robots to make use of the proposed framework.