基于机器视觉的镜片表面缺陷检测系统研发

Reaearch of Lens Surface Detect System Based on Matlab

针对传统镜片企业人工抽样检测的弊端,进行机器视觉测量系统的搭建,以图像处理技术为基础,利用matlab进行缺陷识别算法的开发,判断是否有缺陷,将数据传送至arduino单片机,通过单片机的处理,通过机械手将缺陷镜片剔除,并在人机可视化界面上显示缺陷,从而实现机器视觉在线检测。该系统性能稳定、能够大幅提高检测效率和精确度,同时还能够减少大量劳动力,降低检测工作对工人眼睛的伤害,产生极大的经济效益和社会效益。

For dealing with the disadvantages of manual sampling inspection of traditional lens enterprises,this thesis is written to in- troduce a mechanical visual measurement system. Based on image processing technologtgy,this system is to tesi lens surface defect by using Matlab defeet recognition algorithm. Data gathered from the system will be conveyed to Arduino MCU. After processing,the defect lens will be selected out by mechanical arms. Meanwhile,the defects are to be shown on the visual interface of computer,and thus making the re- al-time detection feasible. The mechanical visual measurement system is of great stability,eficieney and accuracy. It also could decrease labor consumption and reduce the harm to workers＇ eyes. It is of high benefits to introduce the mechanical visual measurement system.