Abstract
This paper proposes a virtual body measurement system by using a Kinect depth camera to extract human body features represented on the 3D body shape. The captured point clouds data are analyzed to extract feature curves and girth measurements of human body shape without pre-marking. The data are processed and aligned to obtain a smooth surface for 3D body shape construction. This method has been successfully tested on human subjects by obtaining the anthropometric data for fitting appropriately-sized clothes. Therefore, the newly developed method can achieve an efficient approach for anthropometric data collection and, in addition, the results can also be visualized through the virtual fitting of real people.