机场视觉移动服务机器人的设计与研究

民航事业迅猛发展,加上疫情突现,机场中迫切需要一种无人的机器人移动服务平台,以减少人力劳动和阻断病毒传播;依据机场环境和功能要求,对机场移动服务平台进行模块化设计与搭建;平台集成RGB-D相机系统Kinect 2.0,采用张正友标定原理确定相机内外参数;设计和实现移动平台的视觉SLAM整体架构,利用ORB算法进行图像特征点的提取与匹配,对邻帧图像进行估计与优化;将平台样机投入实验场地,设定粒子数为50,最小匹配得分为30,构建分辨率为每像素0.02m的二维栅格地图和三维点云图;该样机为机场提供一种以视觉传感器为基础的移动服务平台,结合实际需求,其他功能模块能快捷地集成于此,发挥各自的服务作用。

Design and Research of Visual Airport Mobile Service Robot

With the rapid development of civil aviation and the epidemic, there is an urgent need for an unmanned robot mobile service platform in the airport to reduce human labor and restrain the spread of virus. The airport mobile service platform is modular designed and built according to the airport environment and functional requirements. The platform is integrated with Kinect 2.0 camera system, which is calibrated by Zhang Zhengyou calibration principle to calculate the camera''s internal and external parameters. The framework of visual SLAM (simultaneous localization and mapping) is designed and implemented. Then ORB (Oriented Fast and Rotated Brief) algorithm is used to extract and match the image feature points. Then, the images of adjacent frames are estimated and optimized. In the experiment, the platform prototype was put into the experimental site to draw the map, including the two-dimensional map and three-dimensional point cloud map. The prototype provides a mobile service platform based on visual sensor for the airport. Combined with the actual demand, other functional modules can be integrated into this platform simply to play their respective service roles.