摘要
针对传统Hough变换的参数空间所需存储量大、运算量大、不利于实现实时监测等缺点,提出了一种基于随机Hough变换的改进Hough变换算法,并将该方法用于工件表面纹理图像的处理,通过对工件表面的二值边缘图像进行改进Hough变换,提取直线段平均长度及直线段与切削速度方向的夹角作为特征参数,实现对刀具磨损状态的判断。算法分析和实验结果表明,该算法所需存储空间少,计算量小;提取的特征参数与刀具磨损状态之间存在密切联系,根据特征参数的变化规律可实现对刀具磨损状态的监测。
In view of traditional Hough transform with some shortcomings,such as parameters space requiring large storage,large amount of calculation and unfavor to real-time monitoring,this paper proposes an improved Hough transform algorithm based on Randomized Hough transform,and this algorithm is applied to process the images of the machined surface,and the improved Hough is used to deal with the edge images,from which the lengths and orientations of line segments are distilled as feature parameters to determine the tool wear condition.The results from the algorithm analysis and experiments indicate that this algorithm requires small storage space and small amount of calculation,and that there is a close connection between the distilled feature parameters and tool wear states.Accordingly,based on the variation law of feature parameters,the tool wear states can be monitored.
出处
《西安理工大学学报》
CAS
北大核心
2010年第1期101-105,共5页
Journal of Xi'an University of Technology
基金
陕西省教育厅基金资助项目(075K335)
