基于Delta机器人的视觉分选系统的研究

发布时间：2018-06-17 22:12

本文选题：Delta机器人 + 视觉分选　；参考：《华侨大学》2017年硕士论文

【摘要】：在传统生产线上,产品分选采用人工的方法进行,不仅效率低而且出错的概率极高。即使一些生产线采用工业机器人进行分选,但机器人的一系列动作和产品的摆放位置都是预先精确设定的,机器人只是完成点到点的动作,生产节拍固定,产品种类单一。为了解决生产线上分选作业高度结构化、柔性差等问题,就需要给机器人配备视觉系统。本文从包装盒分选的实际需求出发,以实现Delta机器人对包装盒盒盖与盒底的区分、定位、盒盖的抓取、盒盖的角度调整、将盒盖安装在盒底上及盒盖上Code128A码识别等功能作为分选任务,进行了以下四个方面的研究。第一、从Delta机器人、人工光源、相机、镜头及视觉分选系统开发平台等五个方面详细分析了视觉分选系统的硬件结构,设计了视觉分选系统的软件结构,完成了基于Delta机器人的包装盒视觉分选系统的软硬件平台的搭建。第二、为了实现基于视觉信息的Delta机器人末端执行器对盒盖的抓取,对相机进行了标定,使用标定得到的数据对图像的畸变进行了校正。分析了Delta机器人正逆运动学,对视觉系统进行了标定,得到了相机坐标系与Delta机器人坐标系之间的平移旋转关系,通过基于目标位置反馈的Delta机器人抓取实验验证了本文设计的视觉分选系统的可行性。第三、为了实现Delta机器人对盒盖与盒底的区分、定位及Code128A码识别等功能,开发了基于极坐标霍夫变换的直线检测工具块、基于形状角与霍夫变换的圆检测工具块及基于改进Harris角点检测的正方形检测工具块。根据最小方差原理,设计了一种用于识别Code128A码的方法。第四、设计了一种通过参考坐标点计算盒盖角度调整值的方法。通过Modbus协议实现了视觉系统和Delta机器人之间的数据传输。结合以上研究,实现了Delta机器人对包装盒的视觉分选功能。多次的实验结果表明,本文设计的视觉分选系统运行稳定可靠,包装盒分选成功率达92%以上,具有一定的实用价值。
[Abstract]:In the traditional production line, product sorting is carried out by artificial method, which is not only inefficient but also has a high probability of error. Even though some production lines are sorted by industrial robots, a series of robot actions and product placement positions are predefined, the robot only completes point-to-point movements, the production beat is fixed, and the product category is single. In order to solve the problems of high structure and poor flexibility in sorting operation on production line, the robot needs to be equipped with visual system. In this paper, according to the actual needs of packing box sorting, the Delta robot can distinguish the box lid from the bottom of the box, locate the box cover, grab the lid, adjust the angle of the box cover. The function of code 128A code recognition on the bottom of the box and on the lid is used as the sorting task, and the following four aspects are studied. First, the hardware structure of visual sorting system is analyzed in detail from five aspects: Delta robot, artificial light source, camera, lens and visual sorting system development platform, and the software structure of visual sorting system is designed. The hardware and software platform of packaging box visual sorting system based on Delta robot is built. Secondly, the camera is calibrated and the distortion of the image is corrected by using the calibrated data in order to realize the Delta robot terminal actuator grabbing the lid based on visual information. The forward and inverse kinematics of Delta robot is analyzed, the vision system is calibrated, and the translation and rotation relationship between camera coordinate system and Delta robot coordinate system is obtained. The feasibility of the visual sorting system designed in this paper is verified by Delta robot grab experiment based on target position feedback. Thirdly, in order to distinguish the cover from the bottom of the box, locate and identify the code of code 128A, a linear detection tool block based on polar coordinate Hough transform is developed. Circle detection tool block based on shape angle and Hough transform and square detection tool block based on improved Harris corner detection. According to the principle of minimum variance, a method for identifying Code128A codes is designed. Fourth, a method for calculating the angle adjustment of the lid by reference coordinate points is designed. The data transmission between vision system and Delta robot is realized by Modbus protocol. Combined with the above research, the visual sorting function of Delta robot to the packing box is realized. The experimental results show that the visual sorting system designed in this paper is stable and reliable, and the packing box sorting success rate is more than 92%, which has certain practical value.
【学位授予单位】：华侨大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TP242

【相似文献】