3D profiling of cloth appearances based on stereo vision method

Cloth has the nature of three-dimensional, flexible and non-rigid, some traditional methods based on 2D imaging and contact-type 3D surface measurement methods can not work for this kind of material. In this paper we present a study of digitalization of three dimensional cloth appearances based on stereo vision method. The 3D stereo system includes a set of mirror group (3D lens) and a digital camera with high resolution. Based on the camera geometry and stereo matching techniques, the depth map of cloth could be calculated so as to evaluate the 3D appearance of fabrics and related objects. The results show it is effective and workable to use this system to study the cloth/garment's 3D deformation or draping profile.