摘要
近年来,基于机器视觉的自动检测与定位技术已成为国内外研究的热点。传统的人工视觉检测法、电学参数检测法存在检测效率低、主观偏差因素大、检测范围受限等缺点,已不能满足生产自动化和高质量的要求。因此,本文研究和开发了一个集光、机、电、自动化为一体的精密机器视觉检测设备,以替代人工解决传统电视生产线中屏幕来料检测及整机检测所存在的检测效率低等问题。文章主要实现了四维可调取像技术、多相机三维运动技术、图像有效区域分快提取技术、图像自适应去燥技术以及基于特征值和经验数据库的缺陷类型识别技术,关键技术的实现提高了设备的检测精度、工作效率以及产品质量的一致性和标准化程度。
In recent years, machine-vision-based automatic detection and positioning technology have become a hot research topic.Traditional artificial vision methods and electrical test parameters methods suffer from low detection efficiency, subjective deviation factors and the limited detection range. Therefore, this paper introduces a precise machine vision detection system with the integration of light, machine, electricity and automation, which can replace the artificial detection of incoming material in thetraditional production line of TV screen and largely improve the detection efficiency of the whole machine. This paper implements the 4D adjustable image capture technology, the 3D motion of multi-cameras technology, effective regional image of block extraction technology, adaptive image of noise elimination technology and flaw recognition based on feature value and experience database.Implementation of key technologies to improve the detection accuracy and working efficiency of system, While improve the consistency and standardization of product quality.
作者
陈超
沈航
韦燕华
陈志列
CHEN Chao;SHEN Hang;WEI Yan-hua;CHEN Zhi-lie(EVOC Intelligent technology co., LTD, Shenzhen, Guangdong 518057, China)
出处
《机电工程技术》
2016年第A01期41-46,共6页
Mechanical & Electrical Engineering Technology
关键词
机器视觉检测
四维可调
三维运动
分块提取
识别
Machine vision detection
4D adjustable
3D motion
block extraction
recognition