Study On Subjective And Objective Evaluation Of Bagging Behavior Of Woven Clothing

With the further improvement of living standard,people's requirements for the quality of clothing are becoming more and more stringent.They not only want the clothing to be aesthetic and comfortable,but also have good shape retention.Bagging is a plastic deformation caused by human body movement at the knee,elbow and other parts of the garment,which seriously affects the appearance and aesthetic of the clothing.The static deformation of constant load or constant bagging height is mostly studied in the existing bagging experiments,which is quite different from wearing bagging behavior.The method to evaluate the bagging performance is usually based on image processing or bagging height prediction based on fabric parameters.However,bagging is a kind of 3D dynamic deformation,it is not comprehensive to evaluate it only by 2D image processing or single index.In view of this situation,the experiment of human body dressing was designed to reproduce the bagging behavior.The 3D point cloud data of fabric were obtained by 3D laser scanner,and the 3D indexes of fabric surface were extracted.2D indexes were extracted by image processing technology.Subjective evaluation were carried out by synthesizing the 2D image of bagging and the 3D surface reconstructed by point cloud.Finally,the relationship among 3D indexes,2D indexes and bagging grade was analyzed.The main work and conclusions are listed as follow:(1)The 3D point cloud data of bagging fabric were obtained by laser scanner.With the help of Mat Lab software,the residual bagging height H_max were extracted by using scatter extremum;the residual bagging volume V was obtained by Simpson's rule;the warp,and weft maximum bagging rate?_a(max),?_b(max)were obtained by polynomial fitting method.This geometric quantification indexes were analyzed with single variable to evaluate bagging grade.The results showed that,the new indexes warp and weft maximum bagging rate are significantly related to bagging grade and can be used to distinguish bagging shape.The order of superiority of single variable in predicting bagging grade is as follows,bagging volume V>warp maximum bagging rate?_a(max)>bagging height H_max>weft maximum bagging rate?_b(max).(2)The feature extraction was performed on the processed data to get surface quantization indexes:meanZ,variance M,kurtosis R_k,average deviation R,roughness?and torsion S.Then univariate analysis was carried out between these indexes and bagging grade.The results showed that,the order of superiority of single variable in predicting bagging grade is as follows,roughness?>meanZ>average deviation R>variance M.(3)The regression models of geometric quantitative index,surface quantitative index and fabric bagging grade were established respectively.The results showed that the fitting effect of stepwise regression is better than ridge regression.The geometric quantitative regression model has 94%goodness of fit and 93.3%accuracy due to the selection of the warp and weft maximum bagging rate.The goodness of fit of the optimal surface quantitative regression model is 92.4%,and the test accuracy is 90%.(4)The accuracy of using 2D index(standard deviation of entropy)to predict bagging grade(76.6%)is lower than 3D index,because 3D laser scanning can better consider the spatial information and shape of the fabric surface than 2D image processing,and is not easily affected by light.