Study Of 3D Human Body Contour Digitization And Dimension Measurement Technique

Human body digitization and dimension measurement technique obtains 3D coordinates of points on human body contour through 3D laser sensor system. By data pretreatment, 3D model reconstruction and model editing we can gain human body digitization surface model and then we can obtain the digital expression of human body characters by defining and measuring human body character dimensions in human system engineering theory. This technique can be widely applied in costume industry, medicine, human system engineering, animated cartoon and entertainment.This text adopts 3D human body digitization data captured by 3D laser scanner that is developed by our lab. Plenty of operations are realized such as data pretreatment, 3D model reconstruction, model editing and human body measurement. The main study and innovation are as follows:1. For point cloud data pretreatment technique, I present that deleting noise point by two algorithms of manual and interconnected domain. Aiming at multi-sensor scan data I use ICP algorithm in multi-sensor data registration. Present an algorithm based on mesh processing to merge multi-sensor data. Use distance method and cube method to realize point data reduction.2. For 3D model reconstruction and model editing technique, realize 3D human body model reconstruction by Gopi algorithm. Facing body triangle mesh model, I use decimation algorithm based on vertex deletion criterion to realize triangle meshes decimation. Taubin algorithm and loop algorithm are respectively applied in mesh smoothing and mesh subdivision.3. For 3D model technique in anthropometry, I analyze the influence for the measurement precision when measuring in different data phase and then present the definitions and measurement results of 21 important isosurfaces'positions and 35 important dimensions on human body. I present the mesh processing method in body contour's perimeter calculation and then compare and estimate it with the convex method in perimeter calculation. Theory analysis and experimental result both indicates that measuring perimeter by mesh processing method after point data registration can obtain the highest veracity.