Research On Medical Rubber Plug SPT-P Defects Detection Algorithm Based On Machine Vision

Medical rubber plug,as a critical component in double tube infusion bag,can avoid cross infection.Defects often happen in the process of production,however traditional detection of SPT-P defects is completed by manpower which with big workload and susceptible to effect by personnel subjective factors,so it is difficult to ensure the efficiency and precision detection.But the machine vision detection can improve the production efficiency and ensure the quality of product,which has great practical value.The defects of medical rubber plug mainly distribute in the top and side regions,in order to solve the problem of different defect detection,this paper presents the top side of the defect detection algorithm and the lateral side of the defect detection algorithm.Top defect detection algorithm is mainly used for air bubbles defect detection and aluminum film defect detection,but lateral defect detection algorithm is mainly used for testing the detection of hair silk,the inner cover and oil.(1)The top defect detection algorithm mainly aimed at air bubble defect detection and aluminum film defect detection.First of all we need to extract large disk ROI area like a bubble in air bubble defect detection,aimed at the problem this paper proposes a new center localization algorithm,the detection speed and center location precision of the algorithm can satisfy the detection of air bubble defect.Then the adaptive threshold segmentation method was used to extract the characteristics of the air bubble image.Finally use the area and the length of the fine as the judgment standard of bubble defect.In aluminum film defect detection first we need to extract aluminum film form small disk ROI area,according to the problem we put forward the localization algorithm of least squares straight line fitting to determine the components of angle,and then determine the position of the aluminum membrane disk.Finally use the area and the circumference of the aluminum film small disk position as the detection standards of aluminum film defect.After the simulation experiment on top of defect detection algorithms found on bubbles,aluminum film detection effect is ideal.(2)Lateral defect detection algorithm is mainly used for hair silk,the inner cover and the oil defect testing.Wool yarn exists in the cylinder cover within the surrounding area,oil defects may exist in any area of the components on the side.So before we judge the defect of the hair silk,the cover and the oil,we need to position the ROI regions of the components,and then locate the defect area.Preliminary positioning,first of all,by comparing the two kinds of target localization algorithm,we choose the fast positioning algorithm.Then we use morphological even the brick mark to the follow-up positioning of the cylindrical area and ship type area.Secondly,we use the device location coordinates to extract the defect of ROI regions.Finally,we use geometry characteristic information area,height and width judgment wool yarn defects,using cover defects in grayscale average judgment,smeary defects using gray-level threshold.On the top defect detection algorithm with lateral defect detection algorithm simulation experiment found that inside the bubble,aluminum film,hair silk,cover and oil defects detection effect is ideal.After the simulation experiment on lateral defect detection algorithm found on fuzz,cover and oil defects detection effect is ideal.