A Fast Hole-filling Strategy of 3D Scanned Human Body

Human body scanning technology has a promising application in many fields such as anthropometric surveys, clothing design. Due to the properties of surface (e.g. parallel to camera, occlude, reflectance) cause holes or gaps on the Human Body scanned(HBS) data (e.g. corst, top of head) which challenged the HBS data in applications. Since scanned data usually got noise points near the hole or on the edge, the result of typical boundary edge based filling-hole strategy is tend to be affected by the noise points. And the muti-times-scanned method of filling hole will encountered with high cost both of time and labor consume. This paper presents a filling hole strategy based on point cloud for HBS data. The algorithm is free from identification and triangulation process which is more simply and effectively with less time costs and the method is immune to affection of the noise points near boundary of the holes in contrast to the typical hole filling algorithms.