On-line Detection System Research About Surface Of Transparent Glass Plate

In industrial production,intelligent mobile screen cutting and re-process will inevitably lead to quality problems such as scratches,chipping,and other foreign matter as a result of technical conditions,the production environment and other human factors.These problems not only affect the consumers of the product but also the optical transparency and consistency of the screen,greatly reducing its mechanical strength and the use of the product.Therefore,we must filter products that don't meet the standards to rework through rigorous testing process.The traditional artificial detecting has low detection efficiency,high labor intensity,the test results can not be digitized and met the business requirements of the rapid production.Hence,Important issue of transparent glass plate production quickly and accurately defects detected in product group by group should be solved.Based on a wide class of defect detection literature,understand the basis of similar intelligent detection equipment on the screen surface scratches,point defects,foreign matter defect as a detection object,the investigation of some key technologies for detecting surface of the product based on machine vision has been thorough studiedTo achieve the glass surface of the screen quickly and accurately online testing requirements,we proposed C/S system architecture,IPC as the host to control,multiple graphics workstations as image acquisition and analysis unit.And a comparative study of a variety of defects Extraction Algorithm,final proposed using processing algorithms based on surface morphology,which has a short time-consuming to extract defect information completely.And then measuring a large number of known defects and determine the type of defect features glass screen Criterion.In order to verify the feasibility of detection schemes,built glass detect platform to carried out related experiments,The results show that: This system is able to adapt to the production line speed of 30mm/s,the false detection rate was 7.86%,2.21% rate of missed detection,the detection accuracy up to 10?m × 10?m.Detection methods and criteria be proposed can accurately and quickly identify defects on glasses,meeting the requirements of the glass screen surface defects.