Research On The Automatic Measurement Of 3D Model Of Human Body

3D human body scanning and automatic size measurement technology is an important research field in computer vision.With the gradual improvement of people's living standard and the rapid increase of aesthetic ability,the custom-made industry has been on the rise.Costume customization has drawn more and more attention as an important part of private customization.In the increasingly high degree of automation today,automatic measurement of body size technology is particularly important.The dissertation is devoted to researching the automatic measurement technology of human body 3D model size for garment customization.To get a high-precision 3D model of the human body,we built a 3D body scanning system composed of a multi-camera matrix.We use the method of structure from motion to estimate the camera's internal and external parameters and reconstruct the human body 3D model with high precision through stereo matching algorithm.We propose a method of cutting the human body model parallel.We use the maximum distance method to complete the division between the head and torso.Then we did a split between the upper extremity and the torso by looking for acromion and axilla.At last we use the method of iterative fitting circle to complete the segmentation between trunk and lower extremity.With the reference of "clothing anthropometric parts and methods" we have identified the feature points to be positioned and use the projection method and the human scale factor method to find the human's feature points.We measured a total of 51 parameters for garment customization.Finally for different gender,different body size of the experimental human body,we use the design of the 3D body scan system to complete the 3D reconstruction.Then we performed the segmentation of the reconstructed model and completed the size measurement.We compare the experiment results with the costume designer's manual measurement and verified the effectiveness of this algorithm.