摘要
针对目前刀具损伤检测系统难以从采集的机床刀具损伤图像中自动识别到刀具损伤位置并精准测量刀具损伤量的难题,提出了一种采用局部阈值分割的刀具损伤视觉检测方法。该方法对采集的刀具图像进行灰度化、滤波降噪、旋转定位校正后,将刀具图像均分为多个小像素块,对每个像素块进行图像分割,获取每个像素块的分割阈值,即局部阈值,以最大的局部阈值为基准,对刀具图像进行整体像素扫描,同时结合形态学操作,实现刀具损伤位置识别,并基于识别到的损伤信息精准测量刀具损伤几何特征。搭建了离线检测试验平台,验证所提方法的有效性。试验结果表明:所提方法能够解决目前难以从刀具损伤图像中自动识别到刀具损伤位置并精准测量刀具损伤量的难题,与现有的局部方差法、自适应阈值法等方法相比,刀具损伤几何特征测量的平均准确率至少提升19%以上,具有较大优势。
A tool damage visual detection method using local threshold segmentation is proposed to solve the problem that the current tool damage detection systems are difficult to automatically identify the location of tool damage and accurately measure the amount of tool damage from the collected machine tool damage images.Firstly,the tool image is divided into several small pixel blocks after graying,filtering,noise reduction,rotating and positioning correction.Each pixel block is segmented to obtain its segmentation threshold,that is,the local threshold.Then based on the largest local threshold,the tool image is scanned for global pixels.At the same time,combined with morphological operation,tool damage location recognition is realized,and the geometric characteristics of tool damage are accurately measured based on the identified damage information.An off-line detection experiment platform is built to verify the effectiveness of the proposed method.Experimental results show that the proposed tool damage visual detection method using local threshold segmentation solves the difficulty in automatically identifying the tool damage location and accurately measuring the amount of tool damage from the tool damage image.Comparisons with the existing local variance method,the adaptive threshold method and other methods show that the proposed method improves the average accuracy of tool damage geometric feature measurement by at least 19%,and has a great advantage.
作者
叶祖坤
李恒
查文彬
何彦
王禹林
YE Zukun;LI Heng;ZHA Wenbin;HE Yan;WANG Yulin(School of Mechanical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China;State Key Laboratory of Mechanical Transmission,Chongqing University,Chongqing 400030,China)
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2021年第4期52-60,共9页
Journal of Xi'an Jiaotong University
基金
国家自然科学基金资助项目(52075267)
重庆市自然科学基金资助项目(cstc2020jcyj-msxm2526)
中央高校基本科研业务费专项资金资助项目(30920010005)。
关键词
刀具损伤
视觉检测
自动识别
局部阈值分割
tool damage
visual detection
automatic identification
local threshold segmentation
