Surface Defect Detection Technology Based On Infrared Fringe Deflection

The 14th Five-Year Plan proposes that the traditional manufacturing industry should develop in the direction of intelligence,which also puts forward new requirements for surface inspection technology.In traditional industrial manufacturing,especially in automobile manufacturing,due to the limitations of science and technology and the complexity of the production environment,manual inspection under the strong light are still used to conduct random inspections on products.The cost of manual inspection is high,the efficiency is low,and manual inspection is performed There is a lot of subjectivity,lack of repeatability and quantification,which can easily lead to false detections and missed detections.Therefore,surface defect detection technology which using optical reflection and combined with machine vision has been paid more and more attention.Fringe deflection technology is a full-field gradient technology,which has very high sensitivity to surface height fluctuations.A necessary prerequisite for the application of fringe deflection technology is that there is at least part of the specular reflection on the surface.However,most of the original surface has high roughness in industrial manufacturing,and it is difficult to produce specular reflection within the scope of visible light.In this paper,we proposed using 0.3 mm diameter of nickel chromium alloy wire heating as light source,and extending the incident light wavelength to infrared wavelengths,which making the rough surfaces in the visible light range can produce the rough surface of high mirror reflectivity,and by using infrared camera to observe the surface morphology of heating wire as the light source of the change of hot streaks.Then we can demodulated the change of the surface height from the fringe variable,and the accuracy of system defect detection can reach 89% at the scanning speed of 2cm/s.This paper focuses on the surface defect detection method based on infrared stripe deflection technology,including four main parts: infrared stripe deflection detection principle analysis,detection system hardware platform design and construction,fringe image processing and defect detection system experimental analysis.The main research work of the thesis includes the following four aspects:1.The principle of infrared stripe deflection detection technology is analyzed.According to the electromagnetic wave reflection model of the rough surface,the relationship between the surface specular reflectance and the wavelength of the incident light and the surface roughness is deduced,and the principle of triangulation is used in the fringe deflection technique to complete the three-dimensional reconstruction of the surface topography of the object.2.Complete the design of the hardware platform of the infrared stripe deflection surface defect detection system.Based on the analysis of the principle of triangulation,two surface defect detection systems are designed,and the main components needed to be selected according to the principle of stripe deflection technology,including cameras,light sources,motion control modules,etc.,are completed Platform construction.3.Stripe image processing.The stability of the stripes is analyzed.First,the stripe image is preprocessed to improve the image quality,and then the center coordinates of the stripe are extracted by the gray barycenter method,and the 3D point cloud data of the object surface is obtained from the parallax image,and it is completed by the Delaunay triangulation method.Reconstruction of 3D point cloud data.4.Experimental analysis of infrared stripe deflection detection system.The advantages and disadvantages of two different light sources in the infrared stripe deflection technology are studied.Then we tested the system with 0.3mm diameter NiCr alloy wire as the light source on the actual production line,and the accuracy of system defect detection reached 89% at a scanning speed of 2cm/s.Finally,we analyzed the three factors that affect the accuracy of the detection system from the hardware point of view.