ESTIMATION OF HUMAN BODY SHAPE PARAMETERS USING MICROSOFT KINECTSENCOR

Abstract

In the paper a human body shape estimation technology based on scan data acquired from sensor controller Microsoft Kinect is described. This device includes an RGB camera and a depth sensor that provides, for each pixel of the image,a distance from the camera focus to the object. A scan session produces a triangulated high-density surface noised with oscillations, isolated fragments and holes. When scanning a human, additional noise comes from garment folds and wrinkles. An algorithm of creating a sparse and regular 3D human body model (avatar) free of these defects, which approximates shape, posture and basic metrics of the scanned body is proposed. This solution finds application in individual clothing industry and computer games, as well.