Research on Adaptive Segmentation Method of 3D Point Cloud Data

Aiming at the problem of low efficiency and accuracy of traditional 3d point cloud data segmentation algorithm when kinect vision sensor collects human point cloud data, an adaptive segmentation algorithm of human skeleton points combined with direct filtering was proposed. The direct filter image segmentation through the human body skeleton information to calculate the center of the human body coordinates as a benchmark set threshold value range, compared with traditional direct filter different image segmentation need to specify a target segmentation dimensions and domain's shortcomings, in this article, through the human body skeleton data combined with direct filter method can achieve the goal of adaptive segmentation