Design And Implementation Of Body Shape Measurement System Based On 3d Reconstruction

With the improvement of economy,more and more people begin to pay attention to health and body shape.Body shape measurements can be made with the WBX body scanner,but the device is too expensive.The body shape measurement can be realized based on the 3D reconstruction technology.This method has the characteristics of low cost and wide usage.A body shape measurement system is designed and developed based on 3D reconstruction.Firstly,the monocular camera was calibrated.Then the calibrated camera was used to collect human body images,and these images were preprocessed.Then SIFT(Scale-Invariant Feature Transform)algorithm was used to extract and match the features,and the mismatched point pairs deleted by using RANSAC(Random Sample Consensus)algorithm.Based on the matched feature points,incremental Sf M(Structure from Motion)algorithm was used to generate sparse point clouds on human body surface.Based on the sparse point cloud,the dense reconstruction of human body surface is carried out by using the patch-based algorithm,and the mesh model of human body surface is finally generated.Then the mesh model was sliced,the intersection points of the slices and the human body surface were sorted,and the outer contour of human body was solved.Use the maximum distance to locate the lower neck boundary;The minimum radius of curvature was used to locate the shoulder peak.The axillary position was located by means of point clustering;Use the shortest midline distance to locate the crotch.According to these key parts and the proportion of human body,7 parameters of human body shape were measured.The experimental results show that the system can reconstruct the 3D human body model accurately and provide the body shape measurement report,which can meet the actual needs of users.