基于二维投影的木门尺寸视觉测量

Vision measurement of wooden door size based on two dimensional projection

目前,机器视觉技术广泛应用于工业生产中,零件尺寸的测量是其中一个典型应用之一。针对木门生产线的特点,提出一种简单的尺寸测量方法,对木门图像进行轮廓提取和提取亚像素级角点,提取所有特征点,运用DBSCAN聚类算法对这些特征点进行分类,找到边缘点和去除背景噪点,通过这些角点来提取门的边缘点像素坐标,根据图像像素坐标和相机标定出的焦距与像元尺寸关系,最后得到两点之间的真实尺寸。由于不需要进行二维平面到三维空间的投影变化,图像处理中不需要背景的阈值分割等操作,该算法软件开销很小。经过企业的木门生产线实际应用,证明该处理算法简单、有效。

At present,machine vision technology is widely used in industrial production,part size measurement is one of the typical applications.In this paper,a simple dimension measurement method is proposed to extract and extract the subpixel corner points of the wooden image contours,and all the feature points are extracted.DBSCAN clustering algorithm is used to classify these feature points to find the edge points and remove The noise of the edge of the door is extracted by these corners,and the true size distance between the two points on the image is obtained according to the relationship between the focal length and the pixel size of the image coordinate pixel and the camera.Since there is no need for two-dimensional plane to three-dimensional space projection changes,image processing does not require background threshold segmentation and other operations,the algorithm software overhead is very small.After the practical application of the wooden door production line,it proves that the processing algorithm is simple and effective.