首页|AprilTag图像识别应用于图根控制测量的方法研究

AprilTag图像识别应用于图根控制测量的方法研究

扫码查看
为促进图像识别和人工智能技术在传统测量行业中的应用,提出非接触式的基于AprilTag图像识别的图根控制测量方法,并进行多组实验分别研究观测距离、相机俯仰角、航向角等对图根控制测量精度的影响.实验表明,在机器人相机的航向角∈α(-7.4°,10.6°)、俯仰角β∈(-8.9°,9.9°)、相机中心与AprilTag图像中心的距离L∈(250,300)mm等3种情况下,AprilTag图像识别综合误差E<0.4,识别效果良好,且机器人相机的X轴、Y轴、Z轴等3个方向的坐标测算误差均可控制在毫米级.本文验证了使用人工智能和图像识别技术进行图根控制测量具备一定可行性.
AprilTag Image Identification Method for Application to Tugen Control Measurement
In order to facilitate image recognition and artificial intelligence technology in the traditional measurement indus-try,a non-touch-based AprilTag image recognition method is proposed,and multi-group experiments are carried to study the effect that the observation distance,camera pitch angle,navigation angle exert on Tugen control measurement accura-cy. from. Experiments show that in the cases of robot cam-era α∈(-7.4°,10.6°),pitch angle β∈(-8.9°,9.9°),the dis-tance L∈(250mm,300mm)between the camera center and the center of the AprilTag image,the integrated error of AprilTag image recognition is E<0.4,which proves that the identification effect is good. The coordinate measurement er-ror of the robot camera's X-axis,Y-axis,Z-axis can be con-trolled in millimeter level. This paper verifies the feasibility of this method.

image recognitionAprilTagthree-dimensional coordinate solvingTugen control measurement

曹泽强、洪琼、陈文宇、周立

展开 >

江苏矿业工程集团有限公司,江苏徐州,221000

江苏海洋大学海洋技术与测绘学院,江苏连云港,222000

图像识别 AprilTag 三维坐标解算 图根控制测量

国家重点研发计划"海洋环境安全保障"重点专项海岸带地理环境监测国家测绘地理信息局重点实验室开放基金资助项目连云港国际科技合作计划项目江苏省研究生科研创新计划

2018YFC1405702BE2017125HZ201907SJCX20_1245

2024

测绘地理信息
武汉大学

测绘地理信息

CSTPCD
影响因子:0.563
ISSN:1007-3817
年,卷(期):2024.49(3)
  • 12