Research on filtering and measurement algorithms based on human point cloud data

Abstract

To obtain the data of noncontact measurement of the human body, the depth camera is used to collect the human body, and the obtained initial data are transformed into the required point cloud data for processing through coordinate transformation, and then the collected three-dimensional point cloud data are preprocessed. The preprocessing includes point cloud downsampling, point cloud filtering, plane segmentation, outlier removal, point cloud surface estimation, and so forth. A new solution for point cloud filtering is proposed, which combines sliding least squares and unification and radius filtering. Compared with the traditional filtering, the effect is smoother, and finally the complete outline of the human body is obtained, and then the human body is measured. The results show that the human body data measured by this scheme is within the range of the relevant standard measurement accuracy.