Visual Inspection System For Sealing Defects Of Piston Main Cylinder

Piston master cylinder is used as the brake booster of the automobile.When it is working,the sealing ability of each chamber is guaranteed by the sealing cup installed in the sealing groove.The main types of defects affecting the sealing ability are as follows: 1)There are sand holes and scratches on the surface of the sealing groove;2)The assembly direction of the sealing cup is wrong,which affects the sealing ability.At present,since the diameter of the master cylinder is small,it is difficult for ordinary light source to illuminate the master cylinder.Also,the cylindrical surface of the master cylinder makes it difficult to image at one time.The main method is manual recognition after side-wall imaging using endoscope naked eye detection method.This method has poor imaging quality and low efficiency.Therefore,a visual inspection system based on LabVIEW and NI Vision is designed to detect defects.The main contents are as follows:(1)Analyzing the working principle of the master cylinder and the image characteristics of sealing defects,putting forward the relevant technical indicators and functional requirements,formulating a visual inspection scheme that can identify sealing defects;(2)Designing hardware detection system.The design includes 360 degree panoramic imaging system,mechanical transmission system,air circuit system and motion control system.Among them,the 360 degree panoramic imaging system completes image acquisition and transmission;The vertical and rotating loading mechanism of the mechanical transmission system respectively controls the displacement feeding of the imaging system and the rotation of the measured part;The gas circuit system completes the master cylinder clamping,and the motion control system drives the system to move.(3)Designing an image processing algorithm for defect detection.By using image processing methods such as filtering and denoising,circle detection,proximity interpolation and Gauss mask,the circle image reflected in conical mirror is restored to a rectangular image,and then by using methods like edge detection and area extraction,the defects caused by sands hole can be detected;The scratch defects detection is completed by using multi-directional template and mathematical morphology processing algorithm;and gray projection is used;The direction of the leather bowl was determined by using gray projection method and double peak characteristic threshold method.(4)Designing the software module based on LabVIEW and NI Vision to realize the functions of parameter setting,motion control,image acquisition and processing,result output and interface display.The test results show that the system can effectively detect defects such as sand holes and scratches.In the direction determining experiment of the sealing cup,only 24 errors were detected while the total number is 2000,and the accuracy rate is about 98.8%,which means the system can meet the functional requirements of the system.