The Research On Parameters Reasoning Of Shirt Garment Size In Customization System

In the actual production, the garment specification is mainly based on height and netbust and it is determined by a certain loose measure that the controlling part adds andlooses, which is simple and convenient to confirm the apparel specification and also is infavor of production. However,the body with the different shape and circumference usingthe same type of shirt pattern, ignores the individual fit needs of consumers.Therefore, thispaper establishes the BP model of the shirt size parameters with individualized patterncharacteristics to satisfy the different body fit requirements of consumers，which regardsthe shirt size parameters related to the human net size as the input vector and the shirtclothing specifications as the output vector. The following is the main research’s contents:(1)The selection of the shirt size parameters reasoning method.The particle swarmalgorithm, the genetic algorithm, and the artificial neural network algorithm are discussedpreliminarily and the BP neural network algorithm is identified as the parameters reasoningmethod of the shirt size parameters inference on the basis of the comparison of theadvantages and disadvantages of the three kinds algorithm.(2)The acquiration of the body measurement data which related to the shirtspecification parameters.700human measuring data has acquired, through the3Dscanner.And the bodymeasurement data was divided into7categories based on the twohuman characteristic parameters: the horizontal sextant angle of the straight line fromback salient point to hip salient point and horizontal line, the horizontal sextant angle of thestraight line from chest salient point to abdominal salient point and horizontal line.(3)The determination and the acquisition experiment of shirt specificationparameters.The shirt specification parameters corresponding7kinds of human body hasbeen determined as output vector of BP model.Through the pattern change、personalizedhuman generation and the fabric physical properties setting, the3D virtual fitting has realized.Rely on the three dimensional fitting technology, the sample of the personalized fitshirt size parameters was obtained.(4)The establishment of the BP modle for shirt specification parameters reasoning.Based on the reasult of BP parameters experiments establishing the BP model based ondifferent human shape.(5)The application of shirt size parameters reasoning in customization customizedplatform.Under the environment of Matlab7.1,the realut of the reasoning is made to beapplied in the customization customized system. The main key technologies of the thirdplatform inclued the calculation of the body feature index based on the human sidephotos、the fuzzy recognition of the human body、the reading of the3D scanning file dataand the encoding of the MTM system interface file.This paper set up a BP reasoning model which take the body net size as the inputparameters,while the shirts specification parameters as the output parameters.Theestablished of the BP modle can be used to solve the system application problem of GerberMTM software,The resolution of this key problem will greatly improve the customizationsystem response speed and custom quality.