【作者】 于乃功； 郑相国； 廖诣深； 冯慧；

【Author】 YU Naigong;ZHENG Xiangguo;LIAO Yishen;FENG Hui;Beijing University of Technology;Beijing Key Laboratory of Computational Intelligence and Intelligent Systems;Digital Community Ministry of Education Engineering Research Center;

【机构】 北京工业大学信息学部； 计算智能与智能系统北京重点实验室； 数字社区教育部工程研究中心；

【摘要】 在灾后危险的环境下,含人员位置信息的情景地图是实现救援、提供补给的重要保障。本文提出了一种含受害人员的灾后情景地图构建方法,主要应用于灾后探测机器人上,完成灾后危险环境地图构建任务及在地图上标定需要营救或补给人员的位置信息等任务。首先以搭载激光雷达,里程计,陀螺仪及深度摄像头的自主移动机器人为主体。通过先验地图设定环境探索路径,实现机器人自主地在环境中进行激光建图与搜索任务。建图过程中,对采集到的图像进行人员检测,根据检测结果得到对应检测目标的像素中心。并结合对应的深度信息和当前机器人的位置,通过三维定位模型推算目标人员在世界坐标系中的位置,并标定在地图上。本文所提的方法可为救援和医疗人员提供受害人员在危险环境下的位置信息,方便可为后续机器人执行其他任务提供导航地图,方便在无人环境下开展求援以及补给等任务,有效避免二次感染或伤亡。更多还原

【Abstract】 In the dangerous environment after the disaster, the scene map containing the location information of personnel is an important guarantee for the realization of rescue and supply. In this paper, a scenario map construction method containing the location information of victims in the post-disaster dangerous environment is proposed, which is mainly applied to the post-disaster detection robot to complete the task of constructing the post-disaster dangerous environment map and calibrating the location information of rescue or supply personnel on the map. It starts with autonomous mobile robots equipped with lidar, odometer, gyroscope and depth camera. Through the prior map setting environment exploration path, the robot can carry out the laser mapping and searching tasks in the environment autonomously. In the process of drawing construction, personnel detect the collected images, and obtain the pixel center of corresponding detection target according to the detection results. Combined with the corresponding depth information and the current position of the robot, the target’s position in the world coordinate system is calculated through the 3 D positioning model and calibrated on the map. The method proposed in this paper can provide rescue and medical personnel with the location information of the victims in a dangerous environment, facilitate navigation maps for subsequent robot to perform other tasks, facilitate the implementation of tasks such as seeking assistance and supplying in an unmanned environment, and effectively avoid secondary infection or casualties.更多还原