Research On The Surface Detection System Of Pile Fabric Based On Machine Vision

Pile fabric has the advantages of light texture,good warmth retention,firmness and wear resistance,etc.,and is widely used in traditional industries such as clothing and home furnishing.So far,the detection methods for the surface quality of pile fabrics are still based on traditional manual inspections.However,manual inspections have low efficiency and unstable accuracy,and cannot meet the development requirements of automation and flexibility in the textile industry.In this thesis,machine vision technology is applied to the textile inspection field,and the surface of the pile fabric is inspected from the aspects of system design,image processing algorithm,three-dimensional model establishment,etc.,in order to improve the automation degree of pile fabric inspection.The main content of the thesis is:(1)According to the process characteristics of the pile fabric and the surface characteristics of the fabric,a machine vision-based detection system for the surface of the pile fabric is designed.The tangential image of the pile fabric is obtained by the principle of light section imaging,and the pile fabric is constructed based on the image subtraction method.Thickness model,from which the thickness image of the pile fabric is obtained.(2)Homomorphic filtering,noise reduction and other preprocessing algorithms are carried out on the image of pile fabric thickness to improve the image quality.The maximum between-class variance method is used to segment the thickness image,and the morphological structural elements are constructed for region filling.On this basis,the Canny operator is used to extract the contour features of the villi thickness.The thickness parameter model of the pile fabric is established according to the edge characteristics of the pile contour to realize the detection of pile thickness.(3)On the basis of the contour edge characteristics of the thickness image of the pile fabric,a three-dimensional reconstruction algorithm for the surface of the pile fabric based on the thickness sequence image is proposed.In order to improve the accuracy of the three-dimensional reconstruction of the surface of the pile fabric,a trilinear interpolation algorithm is used.Interpolation operation.On the basis of the three-dimensional reconstruction model,the surface undulation parameters of the pile fabric based on the three-dimensional model are established to realize the detection of the surface state of the pile fabric.(4)An experimental platform for the surface inspection system of pile fabric based on machine vision was built,and the vision system was calibrated.The pile fabrics of two different processes of pile and cut pile are used as samples for systematic experimental verification.The experimental results show that the system in this thesis can objectively quantify the surface parameters of pile fabrics of different processes and detect the error of pile fabrics.It is less than 0.2008 mm,and the thickness detection error of the cut pile fabric is less than 0.1039 mm,which meets the design requirements.