The System Of Four-way Valve Automatic Brazing Inspection Based On Machine Vision

With the popularity of air conditioning,the demanding of four-way valve brazing has increased gradually.However,in our country,most of enterprises are still depending on hand soldering.But this method exists insufficient,like high labor intensity and low efficiency.Some companies rely on imported equipment,but it has disadvantages of poor working condition,low production efficiency and unstable brazing quality.Machine vision technology has the advantages of non-contact,high precision and rapidity.Industry leaders begin turning their attention to machine vision-based test automation and assembly automation,after solving the problem of production automation.Machine vision technology has become one of interest topics in industry.To improve the speed and precision of four-way valve brazing,an automatic brazing inspection system is proposed.Based on machine vision to locate the four-way valve brazing coordinate accurately.The system includes two parts: hardware system and software system.The hardware system designs the assembly line device.The device is divided into 4 stations,brazing by mechanical arm,150 four-way valves can be detected per hour.The system is using 200 W pixel camera,ring lights and near-infrared backlight to light,and to detect the center coordinates and relative height of the four-way valve automatically.The software system includes calibration of camera and location of center and height of four-way valve.Camera calibration method based on planar checkerboard calibration board,program the calibration's program according to 48mmx48 mm calibration board.Camera calibration is required before experiment.Four-way valve copper pipe's location is divided into two steps: center extraction and relative height detection,mainly through gray algorithm,smooth filter,threshold segmentation,contour based shape matching and projection algorithm respectively to achieve the space coordinate information of four-way valve circle in space.Based on the theoretical analysis,an experimental platform is designed and built.The system detection software and camera calibration experiment has been written based on the OpenCV class library and Visual C++.The experimental results are compared and analyzed,the measurement error is less than 0.5mm.Also analyze the optical error of camera in experiment,and its magnitude is negligible.The results meet the design requirements,the software accuracy is up to 99%.So the four-way valve brazing inspection designed in this dissertation has the advantages of high speed,high accuracy,high efficiency stable system operation and so on.In industrial it has good application and broad development prospects.