Surface Quality Inspection Of Uv Resin Of Art Work Surface Based On Improved LBPV Algorithm

With the popularity of 3D printing technology,some countries even launched the App to provide customer-tailored jewelry and provide art work 3D printing combining with the Internet trend.Besides,the rate of 3D printing products is getting faster and faster,and more and more people are willing to send their own 3D printing gifts for their family and friends in important festivals.Nowadays,the commonest and most extensive 3D printing technology is SLA in which the material is generally UV resin.At present,there is no specific research on image detection of the surface quality of 3D art work domestic and abroad.Futhermore,for the poor manual recognition rate,poor stability,poor accuracy and other shortcomings,the project is proposed for the UV resin quality detection of the art work surface.This thesis aimed at quality detection of UV resin of art works that were produced by SLA technology.While with the absence of national and international standards,this thesis concluded the causes of defects from manufacturing technology and made a reasonable assumption.With the Omron integrated camera as the image extraction equipment,it handled some preprocess as the image is denoised,enhanced,and the region of interest is extracted.Furthermore,three improved algorithms based on Local Binary Patterns Variance and LBPV were proposed as "LBPV + algorithm","fusion Hue LBPV algorithm" and "fusion Hue LBPV + algorithm"."LBPV +algorithm"mainly used the method that removing the center pixels."fusion Hue LBPV algorithm"combined the quantized Hue feature on the basis of LBPV.Therefore,"fusion Hue LBPV + algorithm"combined the two of the former advantages.Then,three improved algorithms were used separately to extract defected feature in order to form the eigenvector and the establish of feature database.And the BP neural network was used to train and predict the sample library and thus defects classification of art work surface could be completed on the matlab.As for the problem that the surface quality of the art work relies no fixed standard,the questionnaire survey method was used to investigate the corresponding crowd.Through the questionnaire,the maximum tolerance value of the defect area and the circular contour factor were obtained so that the quality evaluation system was developed and the Matlab was used to simulate.Finally,the matlab program could output the defect type,defect area,defect area ratio,circular contour factor value and the eligibility of the quality.The experiment showed that the improved three kinds of LBPV algorithm had a good effect in solving the quality detection of photosensitive resin of art works.The quality assessment system is effective.Besides,at the end of the thesis,a quality detection model was put forward to solve the surface quality inspection of UV resin,which provided a reference for the quality judgment of art works.