摘要
由于切叠机叠片之前需要对极片进行纠偏准备,为提高极片定位的精准度与速度,提出一种基于梯度方向改进K-means聚类算法,结合局部优化随机抽样一致性(random sampling consistency,RANSAC)算法获取定位坐标。首先,对图片进行降噪等预处理减少外界干扰;基于Canny算子对图像边缘检测,采用多项式插值的方法得到矩形图像直角附近的亚像素级坐标;之后利用改进的K-means算法对边缘亚像素点进行聚类,实现边缘点的划分;结合改进的RANSAC算法拟合出两条相交的直线;最后提取4个直线交点,即图像4个直角拐点,以此确定极片短边中点。实验表明该算法所得中点的波动幅度较小,具有较高的鲁棒性,且定位速度较快,满足工程需求。
In order to improve the accuracy and speed of pole piece positioning,this paper proposes an improved K-means clustering algorithm based on gradient direction,and combines the local optimized random sampling consistency(RANSAC)algorithm to obtain the positioning coordinates.Firstly,the image is pre-processed such as noise reduction to reduce external interference.Based on the image edge detection of the Canny operator,the polynomial interpolation method is used to obtain the sub-pixel coordinates near the right angle of the rectangular image.Then,the improved K-means algorithm was used to cluster the edge sub-pixels to realize the division of the edge points.Combined with the improved RANSAC algorithm,two intersecting straight lines were fitted.Finally,four straight line intersection points were extracted,that is,four right-angle inflection points of the image,so as to determine the midpoint of the short side of the pole piece.Experiments show that the fluctuation amplitude of the midpoint obtained by the algorithm is small,the robustness is high,and the positioning speed is fast,which meets the engineering requirements.
作者
任永强
胡长路
臧昌禹
REN Yongqiang;HU Changlu;ZANG Changyu(School of Mechanical Engineering,Hefei University of Technology,Hefei 230009,China)
出处
《组合机床与自动化加工技术》
北大核心
2025年第5期33-36,43,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
安徽省重点研究和开发项目(202304a05020079)。
