摘要
针对当前国内外刀具磨损检测的缺点和存在的问题,设计了基于机器视觉的刀具磨损检测方案,分析了刀具磨损检测的原理和识别过程,并结合图像处理的方法,采用自适应中值滤波对刀具图像进行平滑去噪,进一步得到刀具的二值化图像,再采用Canny边缘检测技术提取刀具轮廓信息.最后提出基于人工神经网络的刀具磨损检测算法.
In order to solve the shortcomings and problems of cutting tool wear at home and abroad, a scheme of the tool wear detection based on machine vision is designed. The principle and identification process of tool wear detection are analyzed. Through combination of image processing, the method of self-adapted median filter is adopted to eliminate noise on tool image. Then a binary image of the tool is got. And the technique of canny edge detection is used to extract the contour of the tool. Finally an artificial neural network algorithm for tool wear detection is proposed.
出处
《东华大学学报（自然科学版）》
CAS
CSCD
北大核心
2012年第5期505-508,518,共5页
Journal of Donghua University(Natural Science)
基金
国家"八六三"重大专项资助项目(2012AA041309)
关键词
刀具磨损
机器视觉
图像处理
边缘检测
神经网络
tool wear
machine vision
image processing
edge detection
neural network
