| 注册
首页|期刊导航|华中科技大学学报(自然科学版)|基于目标检测的智能钢筋绑扎机器人研究

基于目标检测的智能钢筋绑扎机器人研究

周燕 柯美翔 文世峰 周诚

华中科技大学学报(自然科学版)2025,Vol.53Issue(8):30-37,8.
华中科技大学学报(自然科学版)2025,Vol.53Issue(8):30-37,8.DOI:10.13245/j.hust.250676

基于目标检测的智能钢筋绑扎机器人研究

Intelligent rebar binding robot based on target detection

周燕 1柯美翔 1文世峰 2周诚1

作者信息

  • 1. 华中科技大学国家数字建造技术创新中心,湖北 武汉 430074
  • 2. 华中科技大学材料成形与模具技术国家重点实验室,湖北 武汉 430074
  • 折叠

摘要

Abstract

To ensure the operational robustness of the steel bar binding robot in practical applications,it is necessary to adjust the distance between the steel bars suitable for the robot in real time and eliminate the interference of non working surface steel bars on the recognition of steel bar intersections,based on the work scenario.A smart steel bar binding robot based on object detection was designed to meet the high adaptability requirements of real-time adjustment for general steel bar binding robots.It can adapt to different steel bar binding scenarios and eliminate interference from non working surface steel bars,achieving autonomous path planning and completing steel bar binding operations.Firstly,the working environment of rebar binding project was analyzed to design the four systems of rebar binding robot:mobile system,sensing system,binding system and control system.In the process of rebar intersection labeling and training recognition,a binocular structured light depth camera was used to filter the rebar mesh in depth to eliminate environmental interference,and the depth-filtered images were collected and labeled with rebar intersections as the data set for deep learning.YOLOv5n and YOLOv5s deep learning algorithms were used to train and test the target work surface dataset.The optimal path planning algorithm was proposed to carry out path planning for recognizing the computed positioning coordinates of rebar intersections.The test results show that the designed intelligent rebar binding robot has a better working effect in practice and can effectively eliminate the interference of non-operating surface layer rebars.

关键词

智能钢筋绑扎机器人/深度学习/目标检测/深度过滤/YOLOv5

Key words

intelligent rebar binding robot/deep learning/target detection/deep filtration/YOLOv5

分类

信息技术与安全科学

引用本文复制引用

周燕,柯美翔,文世峰,周诚..基于目标检测的智能钢筋绑扎机器人研究[J].华中科技大学学报(自然科学版),2025,53(8):30-37,8.

华中科技大学学报(自然科学版)

OA北大核心

1671-4512

访问量0
|
下载量0
段落导航相关论文