摘要
刀具参数的机器视觉测量是高精度机械加工必不可少的步骤,高效的刀具图像处理能极大的提高机械加工的效率.本文利用计算机、CCD相机及可调LED光源构建了图像采集系统,并在VS 2010平台上用C#语言编写了Canny算子边缘轮廓提取算法.对Canny算子的时间复杂度进行分析,通过对比几种经典的阈值分割法的运行时间,选取运行时间最少的Otsu法来提取刀具边缘及附近像素点以减少Canny算子的计算量.使用Query Performance Counter计时器对Canny算子各模块计时,Canny算子运算时间从原来的将近300ms缩短到改进后的100ms.这种改进极大地提高了刀具图像的处理速度,且能广泛地被应用在其他机器视觉项目当中.
Machine vision measurement of tool parameters is essential for high precision machinery processing, Efficient tool image processing can greatly improve the machining efficiency. An Image Collec- tion System is constructed with a computer, a CCD camera and LED source, meanwhile the Canny operator system is compiled by C# on VS 2010 platform based on machine vision. The Canny operator's time complexity is analyzed and the running time of several classic threshold segmentation methods are com- pared. The least running time of the Otsu method is selected to extract the edge of the cutter and the nearby pixels to reduce the amount of calculation of Canny operator. Query Performance Counter timer is used for timing of each module Canny operator. The computation time of the improved Canny operator is reduced from the original 300 ms to nearly 100 ms. This improvement idea greatly improves the speed of image processing tool, and can be widely used in other machine vision.
出处
《测试技术学报》
2015年第2期132-136,共5页
Journal of Test and Measurement Technology
基金
山西省研究生优秀创新项目(20133095)