www.gusucode.com > 基于MATLAB图像处理的弯管检测方法 > 基于MATLAB图像处理的弯管检测方法\文件说明.txt

    为了提高刚性弯管角度与尺寸检测的精度与效率,弥补传统测量方法与工具的不足,提出了一种基于机器视觉的非接触式弯管检测方法。该方法通过预先设定参照工件与待测弯管,利用CCD机摄取两者含有尺寸信息的特征图像,借助MATLAB的图像处理技术提取两者的坐标数据,求解图像上参照工件的尺寸缩放倍数,进行弯管尺寸值的求解;弯管的角度与弯角半径分别采用最小二乘法线性拟合与霍夫变换拟合圆求解。实验表明:本方法检测精度优于传统检测结果;其操作简单、高效,适用于尺寸不大的各类弯管检测。 In order to improve the accuracy and efficiency of the rigidity and size of rigid elbow and make up the deficiency of traditional measurement methods and tools, a non - contact elbow detection method based on machine vision is proposed. The method can be used to extract the coordinate data of the two by using the image processing technology of MATLAB, and the size scaling of the reference workpiece on the image is solved by preliminarily setting the reference workpiece and the measured elbow. The angle of the elbow and the radius of the bend are calculated by the least squares linear fitting and the Hough transform fitting. The experimental results show that the method is superior to the traditional test results. Its operation is simple and efficient, and it is suitable for all kinds of elbow testing.