基于机器视觉的臂架式起重机极限承载性能检测方法  

Machine Vision Based Inspection Method for Ultimate Bearing Capacity of Jib Crane

作　　者：曹旭阳[1] 李强 张则宝 唐旭扬 CAO Xu-yang;LI Qiang;ZHANG Ze-bao;TANG Xu-yang(School of Mechanical Engineering,Dalian University of Technology,Dalian 116023,China)

机构地区：[1]大连理工大学机械工程学院,辽宁大连116023

出　　处：《仪表技术与传感器》2023年第4期113-117,共5页Instrument Technique and Sensor

基　　金：国家自然科学基金项目(52275088)。

摘　　要：针对传统的臂架式起重机承载性能检测试验的不足,提出了一种新的检测方法,通过吊载一辆随动车体来代替吊载的砝码。在新方法中,设计了一种基于圆形特征的单目视觉定位方法,识别特定的几何特征标记来获取随动车体像素坐标系下的坐标,为随动车体在地面上的运动提供方位,对承载性能检测环境进行了分析研究,设计了一种自适应阈值和全局色调映射的图像分割方法,适应户外光照变化大和亮度分布不均的情况,以及应用canny边缘检测识别特征的轮廓。实验验证该图像分割方法对多变的光照条件具有较强的鲁棒性,随动车体模型能够较好地跟随定位装置移动,实现精确定位,新的检测方法可行性高。Aiming at the shortcomings of the jib crane s bearing performance testing test,a new testing method was proposed,which replaced the hoisted weight by hoisting a follow-up trolley.In the new method,a monocular vision positioning method based on circular features was designed.This method identified the specific geometric feature markers to obtain the position of follow-up trolley.The position provided orientation for the movement of the follow-up trolley on the ground.In view of the analysis and research of bearing performance detection environment,an image segmentation method based on adaptive threshold and global tone mapping was designed to adapt to the situation of large outdoor illumination changes and uneven brightness distribution.The contour of the features was identified through canny edge detection.Experiments show that the image segmentation method has strong robustness to changing lighting conditions,and the follow-up trolley model can follow the positioning device well to achieve accurate positioning.The new detection method is highly feasible.

关 键 词：臂架式起重机 承载性能检测 机器视觉 全局色调映射 形状识别 

分 类 号：TP23[自动化与计算机技术—检测技术与自动化装置]