Authors

Abstract

In order to solve the problem of autonomous movement of intelligent picking robot, a method for big data analysis is proposed. The data collected by laser ranging sensor, CCD camera and electronic compass are analyzed to determine the current robot position and heading. The walking route for the robot is preset of the orchard. The sign for the digital tube is installed, so the coordinates for the sign are known. Electronic compass is used to determine the heading angle of robot. The CCD camera captures the sign image, and adopts the methods of image graying, image segmentation, image corrosion and image thinning to extract the digital tube image data on the sign. Fuzzy control method is used to identify digital tube numbers on the ground sign, and CCD camera scans whether there are obstacles in front of the road at the same time. The laser ranging sensor completes two tasks: ① detecting the sign distance and direction angle of the distance, and calculating the current position and heading angle of the robot through the coordinates of the sign; (2) detect the distance from obstacles, and start the obstacle avoidance system when it is less than 1m. The path planning system of picking robot based on big data analysis relies on a variety of sensors, and has a strong ability to perceive orchard environment, and it has high ability of path planning.

Abstract in Chinese

为了解决智能采摘机器人自主运动问题，一种面向大数据分析的方法被提出。对激光测距传感器、CCD摄像机和电子罗盘等传感器采集的数据进行分析，确定当前机器人位置与航向。在果园中预先设定机器人行走路线，安装数码管地标牌，地标牌坐标已知;电子罗盘用于确定机器人航向角;CCD摄像头拍摄地标牌图像，采用图像灰度化、图像分割、图像腐蚀及图像细化的方法提取表示牌上数码管图像数据。采用模糊控制方法识别地标牌上数码管数字，CCD摄像头同时扫描行进前方是否有障碍物。激光测距传感器完成两个任务:①检测距离地标牌距离和方向角，通过标识牌坐标计算机器人当前位置与航向角;②检测与障碍物间距离，当小于1m时启动避障系统。基于大数据分析的采摘机器人路径规划系统依托多种传感器，对果园环境感知能力强，具有较高的路径规划能力。