Research on 3D attributes extraction from cloth 2D picture

The cloth 2D contour animations are capable of being realized by both of the curve fitting method and the cftool box in Matlatb software platform. The comparison of these animations will build solid foundation of the cloth later 3D attributes extraction and simulation. Secondly, several pictures 3D attributes extraction is considered. When 3D attributes extracted from a series of cloth 2D images, key points should be sampled in their corresponding gray images. The z-coordinates values of such key points may be defined by their individual gray values. For each cloth 2D image, its 3D geometrical features can be analyzed and extracted with mathematical modeling algorithms. Similar to the cloth 2D contour animations, the objective curves here may also be fit by two kinds of techniques, such as an interpolated fitting method and the Sftool box. Compared with the two corresponding results, the better is selected as the final 3D simulation result. According to the coordinate information of the sampled feature points, and using a geometric method of plane, the capture position angles relationship of the continuously given cloth images may be calculated, with which the final capture angle is achieved.