The paper present a software application used body shape recognition. The 3D human body scanned can be used for avatars or parameterized models. The parametrized manequins can be useful for garment mass customization. It can make 3D avatars that can be used in the video games industry. The classification method is based on clustering algorithm for body shape recognition. The traditional 3D scan data processing software improve the optimizing polygon mesh models and generate high quality nurbs surfaces from 3D scan data. The human body captured from the Vitus 3D body scanner is not the perfect object one - it is just an opened mesh, because the scanning process areas can occur "noises" which are actually areas where cameras can not take picture because of barriers (arms, legs, hair, etc.). To have a validated body for use in the shape classification is necessary to use the body surface retriangulation process of filling holes that appear in the body mesh surface. It can be use the global remesh to retriangulate the faces and improve the quality of human body object faces. This validated body can offer numerous anthropometric indicators that can be used as parameters in the classification of body shape.
