Research On Objective Evaluation Of Garment Sewing Smoothness Grade

With the increase of income level, life is getting better and better, the quality of lifehave higher pursuit. Clothing as the necessities of life, people also to the qualityrequirement is increasing day by day, and the appearance of the clothing quality is animportant role in the quality of clothing, received a lot of attention. Sewing flatness, as animportant factor of the look of the garment, is becoming more and more attention by theresearchers. Currently, sewing flatness evaluation is mainly against the United StatesAATCC standard-88-b subjective judgment, is the overall quality of garment sewing, isqualitative evaluation method, this method has been widely recognized and used, but isdifficult to avoid the influence of human factors and the environment, the evaluationresults is not stable, there is a big limitation.According AATCC-88B standard8-bit grayscale image, and analyze the grayscaledistribution characteristics, and to extract the feature parameters, establish the sewingflatness objective evaluation model based on BP neural network. In addition, In order toavoid the influence of the printing design for gray image analysis, this paper introducesdouble CCD scanning flatness objective evaluation of the sewing, using binocular CCDscanning technology for ordinary sewing sample image, at the same time, by about twoCCD cameras, will be taken from different Angle of two image registration, sewing on thesamples of3d information, at the same time to avoid color and pattern for the interferenceof the information of the flatness. AATCC-88B in order to sewing smoothness ratingstandard sample as consistent evaluation, will be sewing smoothness of binocular CCD toobtain3d information into two-dimensional gray image analysis. Based on the analysis ofthe Z axis information on3d image, the relationship established between height and grayscale, using VB to design algorithms, the3d image into a two-dimensional gray image, the3d problem is converted into a two-dimensional problem, and then with the help of Matlabsoftware, analysis of two-dimensional image grayscale distribution characteristics, extract the characteristic parameters evaluation seam smoothness. Finally used to extract thecharacteristic parameters of sample test the prediction precision of BP neural networkmodel, implementation for objective evaluation of sewing sample sewing fabricsmoothness and accuracy have been carried out to verify the proposed evaluation model.The actual output and desired output through a network of regression analysis, thecorrelation coefficient is0.93, the evaluation model has good prediction ability. Finally,this paper on the basis of theoretical research, with the aid of VB and MATLABprogramming, has developed a man-machine interactive level of garment sewing flatnessobjective evaluation system. The system can classified management fabric specificationand performance of sewing sample, complete subjective and objective evaluation of seamsmoothness of the sewing sample, at the same time calculation error of subjective andobjective evaluation. The system has friendly interface, high degree of visualization, fastand convenient to data processing, the advantages of simple operation.In this paper, the sewing flatness objective quantitative evaluation research, avoid theambiguity of the sewing appearance quality evaluation, for the quality of fabric or garmentsewing to predict or evaluation provides a effective method. Provide a reference forclothing design, processing and manufacturing.