Object Motion Characteristic Reconstruction Method Based on Random Sampling Consistency

Abstract

The invention discloses a method for reconstructing object motion characteristics based on random sampling consistency. Motion trajectory; perform spatiotemporal clustering on sparse motion trajectories to obtain a series of motion trajectory combinations; calculate the relative motion trajectories of each trajectory combination and other motion trajectories; based on each group of relative motion trajectories, random sampling consistency algorithm to fit the motion characteristics; use the fitted motion characteristics to filter the corresponding motion trajectory combinations, and merge the trajectories whose motion characteristics are similar to a certain threshold; and then reconstruct the motion characteristics of all joints of the entire joint object. By adopting the technical scheme of the present invention, the motion characteristics of the object can be efficiently and effectively extracted from the low-quality three-dimensional scanning data.

Description

technical field [0001] The invention relates to the field of reconstruction using a 3D scanner, in particular to a method for reconstructing object motion characteristics based on random sampling consistency. Background technique [0002] With the development of 3D scanners and 3D scanning technology, it is not difficult to use 3D scanners to scan and reconstruct real-life objects or even scenes. With the unremitting efforts of researchers in many related fields, more and more advanced scanning and reconstruction methods have been proposed. Now people can not only scan and reconstruct static scenes, but even scan and reconstruct dynamic scenes. Whether it is the reconstruction of a static scene or a dynamic scene, in the final analysis it is the reconstruction of the geometric features of the object, and an object not only contains geometric features but also contains motion mechanisms. Compared with the geometric characteristics of the object, the motion characteristics of...

AI Technical Summary

Problems solved by technology

[0003] However, it is very challenging to extract the motion characteristics of the object based on the dynamic 3D data of the object surface captured by the 3D scanner

The main reasons are: 1. Due to the limitations of the resolution and frame rate of the 3D scanner itself, the scanned 3D data points are often sparse and have many noise points, which are incoherent in time

2. The scanned object or its parts may move rapidly during the scanning process. Due to the ghost effect, the data of the moving part of the object often has many abnormal data points

3. Objects in motion may cause data loss due to self-occlusion

However, the time overhead for geometric reconstruction of objects is too large. This method tightly binds geometric reconstruction and motion characteristic reconstruction, which is obviously inefficient.

PEKELNY (PEKELNY, Y., AND GOTSMAN, C.2008. Articulated object reconstruction and markerless motion capture from depth video. Computer Graphics Forum (Special Issue of Eurographics) 27, 2, 399–408.) proposed a template-based method to Carry out geometric reconstruction and motion analysis, however, this method requires the user to manually segment the first frame of data in order to automatically complete the follow-up work

Obviously, it would be tedious to manually interact with each object to be scanned

Images