The technology of 3D garment virtual sewing has been developed for several decades, we proposed an approach to verify the virtual garments sewing effect. Here we took a sleeveless dress for our example, and the patterns of the dresses are made by expertise garment CAD software. The dresses patterns are used for two options. The first is for actual cutting and sewing. The second is for virtual sewing. The actual sleeveless dress is made by woolen fabric for our experiment. The parameters of the fabric mechanical properties in virtual sewing should be selected the same with the actual fabric. The fabric properties in this paper are mainly referred to shear property, tensile property and bearing property. We observe the surface shape of the finished dresses, compare the differences between the true and virtual sewing, and find the modified approach. Gray scale transformation and binary image processing are exploited for appearance comparison between real and simulated clothes sewn by the same 2D patterns. Folds and silhouettes comparison are main two features in our proposed approach. The sum of the squared differences is calculated for folds comparison, and the mismatches pixels are computed for silhouettes comparison. The aim of our research is proposed for forecasting the ready-made clothes which are made by given patterns, and check whether it is meet the requirement of the clients or not.
Average efficiency of data structures for binary image processing
