语义地图导航的智能化服务机器人研究

基于语义地图导航的智能化服务机器人系统,以搭载深度相机和二维激光雷达的差分轮智能车的形式呈现。系统以树莓派作为主控制器,利用OpenCR控制电机驱动和辅助驾驶传感器,通过ROS机器人操作系统将机器人控制器、智能信息处理模块和各类传感器融合成一个有机整体。所构建的机器人系统在服务过程中,通过感知周围环境、收集数据完成静态、动态物体的辨识,然后结合二维地图数据完成目标检测和语义地图的构建。为了提升系统智能化程度,配合深度推理神经计算棒,并结合Mobile Net网络架构下的SSD目标实时检测算法,提高了视觉感知的实时程度以及准确性。

Research on Intelligent Service Robot Based on Semantic Map Navigation

The intelligent service robot system based on semantic map navigation is presented in the form of a differential wheel smart car equipped with depth camera and two-dimensional lidar.The system uses Raspberry Pi as the main controller,with OpenCR to control the motor drive and auxiliary sensors.The robot controller,intelligent information processing module and various sensors are integrated into a completed equipment by using the ROS robot operating system.In the service process,the constructed robot system completes the identification of static and dynamic objects by sensing the surrounding environment and collecting data,then combines the two-dimensional map data to handle the target detection and the construction of the semantic map.In order to improve the system intelligence,in conjunction with the deep inference neural computing stick,combined with the SSD target real-time detection algorithm under the Mobile-Net network architecture,the real-time detecting degree and accuracy of visual perception are improved.