快速Curvelet变换在机器视觉系统图象边缘增强中的应用
Application of Fast Curvelet Transform in the Image’s Edge Enhancement of Machine Vision System
作者:廖世鹏(四川大学 测量与控制工程系,四川 成都 610065);廖俊必(四川大学 测量与控制工程系,四川 成都 610065);李彬彬(四川大学 测量与控制工程系,四川 成都 610065)
Author:(School of Measurment and Control ,Sichuan Univ., Chengdu 610065,China);(School of Measurment and Control ,Sichuan Univ., Chengdu 610065,China);(School of Measurment and Control ,Sichuan Univ., Chengdu 610065,China)
收稿日期:2007-09-03 年卷(期)页码:2008,40(3):168-172
期刊名称:工程科学与技术
Journal Name:Advanced Engineering Sciences
关键字:干簧管;机器视觉系统;快速曲波变换;IPP函数;稳定性
Key words:reedswitche;machine vision system;fast Curvelet transform;IPP functions;stability
基金项目:国防科技基础研究基金资助项目
中文摘要
机器视觉系统在工业中的应用日益广泛,而在应用中要求满足稳定性和实时性的要求。讨论了在干簧管装配仪工作过程中为了使其机器视觉系统能够稳定、实时地监测干簧管的几何参数,应用Curvelet变换进行稳定的图像边缘增强并使用IPP函数进行优化以提高速度,最终使该系统在工业应用中能每12 ms处理一帧图像,其重复精度可达4 μm。经实践证明,应用该方法使干簧管装配仪机器视觉系统达到了稳定、实时监测几何参数的要求。
英文摘要
Machine vision system must meet the requirements of real-time and stability. An effective method was proposed and applied to the machine vision system of reedswitche madding device to monitor the geometric parameters. The edge of the image is enhanced by the Curvelet transform for stability and the program is optimized by IPP functions for high speed.This system can process an image every 12 ms and its repeat accuracy is 4 μm. The application showed prove that the machine vision system of the reedswitche madding device can monitor the geometric parameters through stable and real-time processing by this method.
【关闭】