基于偏振图像处理的塔机位姿估计方法  

Research on Tower Crane Pose Estimation Based on Polarization Image Processing

在线阅读下载全文

作  者:朱达荣[1] 王强 龚雪[1,2] ZHU Darong;WANG Qiang;GONG Xue(Key Laboratory of Construction Machinery Fault Diagnosis and Early Warning Technology,Anhui Jianzhu University,Hefei 230601,China;School of Mechanical and Electrical Engineering,Anhui Jianzhu University,Hefei 230601,China)

机构地区:[1]安徽建筑大学建筑机械故障诊断与预警重点实验室,安徽合肥230601 [2]安徽建筑大学机械与电气工程学院,安徽合肥230601

出  处:《安徽建筑大学学报》2024年第6期42-50,共9页Journal of Anhui Jianzhu University

基  金:安徽省自然科学基金项目(2008085UD09);安徽省住房城乡建设科学技术计划项目(2022-YF016)。

摘  要:实时获取塔机位姿有助于预防塔机安全事故。将相机固定在塔机顶端,可通过求取相机位姿来估计塔机位姿。但在光照变化及较大背景噪声等复杂环境下,基于特征点的相机位姿估计存在精度较低的问题。为此,本文提出了一种基于偏振图像处理的塔机位姿估计方法。首先使用张正友标定法求得相机内参,然后设计靶标并使用连接点法实现图像去畸变,最后通过偏振图像估计塔机位姿。实验结果证明,拍摄距离为30 m时,在不同光照和视角下的偏振图像比光强图像精度更高,位姿平均误差为1.34%。The real-time acquisition of tower crane pose can decrease the risk of tower crane safety accidents.With a camera fixing on the top of the tower crane,the pose of the tower crane can be estimated by calculating the pose of the camera.However,in complex environments such as lighting variations and large background noises,feature point-based camera position estimation suffers from low accuracy.To solve this problem,a method of tower crane pose estimation based on polarization image processing is proposed.The camera's internal parameters are obtained by using Zhang Zhengyou's calibration method,and the image of a designed target is de-distorted by using the connection point method.Hence,the pose of tower crane can be obtained by the polarization image.The experimental results show that in a 30-m shooting distance,the polarization images of the scene,although with different illumination and viewing angles,has higher accuracy than the light intensity images,and the average position error is 1.34%.

关 键 词:塔机位姿估计 偏振成像 光照变化 视角变化 

分 类 号:TP391[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象