基于K均值聚类的金属表面光条中心亚像素提取方法  

Sub-pixel Extraction Method for Light Stripe Center on Metal Surface Based on K-means Clustering

作　　者：金振中 JIN Zhenzhong(School of Mechanical Engineering,Nanjing Institute of Technology,Nanjing 211167,China)

机构地区：[1]南京工程学院机械工程学院,江苏南京211167

出　　处：《现代信息科技》2025年第5期115-118,124,共5页Modern Information Technology

摘　　要：线结构光视觉传感已广泛应用于焊缝检测与跟踪领域,针对焊缝检测过程中金属表面反光干扰的问题,提出一种基于K均值聚类的光条中心亚像素提取方法。首先对采集的结构光条纹图像进行中值滤波处理去除高频噪声,再通过K均值聚类算法分离光条与背景,避免背景漫反射光斑的干扰;对完成聚类分割的光条图像使用开运算处理,平滑光条边缘并去除毛刺;最后使用自适应二维灰度重心法完成对光条亚像素中心的提取。所提取中心线到中心点集最小二乘法拟合直线距离标准差为0.1570像素,程序运行时间为0.2518 s。Line structured light vision sensing is widely used in the field of weld detection and tracking.To address the problem of reflective interference on the metal surface during weld detection,a sub-pixel extraction method for the light stripe center based on K-means clustering is proposed.Firstly,median filtering is applied to process the collected structured light stripe images to remove high-frequency noise.Then,the K-means clustering algorithm is used to separate the light stripe from the background,avoiding the interference of background diffuse light spots.The opening operation is performed on the light stripe images after clustering segmentation to smooth the edges of the light stripe and remove the burrs.Finally,the adaptive twodimensional gray-scale center of gravity method is used to extract the sub-pixel center of the light stripe.The standard deviation of the linear distance between the extracted centerline and the line fitted by the least squares method for the set of center points is 0.1570 pixels,and the program running time is 0.2518 s.

关 键 词：线结构光 中心提取 图像聚类分割 灰度重心法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]