基于双目视觉图像识别的输煤机皮带跑偏在线识别方法
An Online Identification Method for Belt Deviation of Coal Conveyer Based on Binocular Vision Image Recognition
赵芳兵 1徐学东2
作者信息
- 1. 甘肃华亭煤电股份有限公司华亭煤矿,甘肃 华亭 744100
- 2. 华亭煤业集团有限责任公司,甘肃 华亭 744100
- 折叠
摘要
受输煤机皮带原始图像质量的影响,导致对其跑偏程度的识别结果误差难以得到有效控制,基于此,提出基于双目视觉图像识别的输煤机皮带跑偏在线识别方法.将Mech-Eye UHP-140工业级3D双目视觉相机作为具体的输煤机皮带视觉图像采集装置,在对其安装后,结合实际的输煤机皮带运行状态以及图像质量要求,对具体的采集阶段参数进行设置;在输煤机皮带跑偏程度识别结果,引入了亥姆霍兹原理,对双目视觉图像的边缘进行识别,根据双目图像边缘交叉程度与皮带中心线之间的距离关系,确定具体的跑偏距离.在测试结果中,设计对于皮带偏移量的识别结果并未受到运行速度的影响,且具体的识别结果误差始终稳定在0.1 mm 以内.
Abstract
Affected by the quality of the original image of coal conveyor belt,it is difficult to effectively control the error of the identification result of its deviation degree.Therefore,this paper proposes an online identification method of coal conveyor belt deviation based on binocular vision image recognition.The Mech-Eye UHP-140 industrial 3D binocular vision camera is used as a specific vision image acquisition device for coal conveyor belt.After its installation,the parameters of specific acquisition stage are set according to the actual running state of coal conveyor belt and image quality requirements.The Helmholtz principle is introduced to identify the edge of binocular vision image,and the specific deviation distance is determined according to the distance between the cross degree of binocular image edge and the center line of belt.In the test results,the identification result of the design for the offset of the belt is not affected by the running speed,and the specific identification error is always stable within 0.1mm.
关键词
双目视觉图像识别/输煤机皮带/跑偏识别/工业级3D双目视觉相机/亥姆霍兹原理/皮带中心线Key words
binocular vision image recognition/coal conveyor belt/deviation recognition/Mech-Eye UHP-140 Industrial 3D binocular vision camera/Helmholtz's principle/belt center line引用本文复制引用
出版年
2024
NETL
NSTL
万方数据