基于机器视觉的尺寸测量研究进展  

Research progress in dimension measurement based on machine vision

作　　者：曹广如[1] 黄石甫 CAO Guangru;HUANG Shifu(Zhuzhou Times New Materials Technology Co.,Ltd.,Zhuzhou,Hunan 412007,China)

机构地区：[1]株洲时代新材料科技股份有限公司,湖南株洲412007

出　　处：《轨道交通材料》2023年第6期29-33,共5页MATERIALS FOR RAIL TRANSPORTATION SYSTEM

摘　　要：基于机器视觉的尺寸测量是近年来在计算机视觉领域中备受关注的一个研究方向。随着人工智能和图像处理技术的快速发展,利用机器视觉进行精确尺寸测量的方法因其准确、高效、非接触的特性,被广泛应用于智能制造、质量控制和自动化系统等领域。常用的机器视觉尺寸测量方法,包括特征提取、边缘检测、模板匹配和三维重建等技术,通过建立相机与被测物的相对坐标,计算转换得到测量数据。同时,基于机器学习和深度学习的尺寸测量方法,通过建立高效准确的尺寸预测模型,实现更灵活准确测量。机器视觉测量从利用设备获取数据的方式的不同,可分为被动测量与主动测量,文章将针对上述方法在两种方式上的应用,对基于机器视觉的尺寸测量技术进行综述,总结了机器视觉测量存在的问题,并对其发展趋势进行了展望。Dimension measurement based on machine vision has been a highly anticipated research direction in the field of computer vision in recent years.With the rapid development of artificial intelligence and image processing technology,machine vision has been widely applied in intelligent manufacturing,quality control and automation systems for accurate dimension measurement due to its precise and efficient non-contact characteristics.The commonly used dimension measurement method based on machine vision,including feature extraction,edge detection,template matching and 3D reconstruction,can calculate and convert the measurement data by establishing relative coordinates between camera and the object to be measured.At the same time,dimension measurement method based on machine learning and deep learning could establish efficient and accurate size prediction models,and therefore achieve more flexible and accurate measurements.Machine vision measurement can be divided into passive measurement and active measurement,depending on the ways of using equipment to obtain data.This paper reviews the application of the above-mentioned methods in two different ways,summarizes current problems in machine vision measurement,and anticipates its development trend.

关 键 词：机器视觉 图像处理 尺寸测量 智能制造 

分 类 号：U270.7[机械工程—车辆工程]