Point cloud three-dimensional model reestablishing method and system

Abstract

The invention relates to a point cloud three-dimensional model reestablishing method which comprises the following steps that a. a point cloud of an object which needs three-dimensional model reestablishing is obtained in a scanning mode; b. the obtained point cloud is subjected to quality analysis; c. for the point cloud obtained after quality analysis, new scanning viewpoints are obtained by computing; d. according to the new scanning viewpoints, scanning is carried out, and according to the point cloud obtained by scanning the new scanning viewpoints, the point cloud of the step a is updated in real time; and e. according to the point cloud obtained after real-time updating, three-dimensional model reestablishing is carried out. The invention further relates to a point cloud three-dimensional model reestablishing system. A three-dimensional model can be reestablished in a full-automatic mode, a high-quality point cloud model is generated, and meanwhile the achieving scheme is simple and high in efficiency.

Description

Technical field [0001] The invention relates to a point cloud three-dimensional model reconstruction method and system. Background technique [0002] Computer three-dimensional models play an important basic role in people's current life (such as games, film and television, virtual reality, etc.) and production (such as 3D printing, reverse engineering, digital city, etc.) activities. Traditional manual 3D modeling is far from meeting people's growing demand. The bottleneck of development is that there is no means of rapid and automatic modeling. Although 3D point cloud scanning technology has been developed for more than 20 years, it takes a lot of time and effort to scan a model and obtain high-quality point cloud data, even if professional users are equipped with professional equipment. [0003] The disadvantages of the prior art are summarized as follows: [0004] First: relying too much on the reconstructed surface of the existing point cloud. The existing technology requires...

AI Technical Summary

Problems solved by technology

Although this method is efficient, it has two major disadvantages: First, in places with rich object details, there may be a small amount of point cloud coverage during the initial scan and the surface can be reconstructed, but more scans from different angles are needed to obtain Enough point cloud to restore the original appearance, while the existing technology assumes that the point cloud data of the part that can reconstruct the surface is good enough, no further scanning is needed, so a lot of detail and sharp features are lost

Second, for models with complex topology (such as human body, trees, or craft ornaments), the surface generated by the existing surface reconstruction method on the incomplete point cloud is likely to be wrong, but the existing technology assumptions have generated surfaces are all correct, so only models with simple shapes can be processed

[0005] Second: Do not analyze the quality of the existing point cloud

The existing automatic scanning technology does not process and analyze the inherent defects of these point cloud data, but simply uses the coverage of the point cloud on the source model as a guide. The automatically scanned point cloud model will inevitably be quite different from the real model.

[0006] Third: The splicing and updating of point clouds is not efficient

Where the model has features, the scan is often incomplete, and there are many noises and noises. Therefore, this method is still not reliable enough without manual assistance.

In addition, the existing technology is to directly add all newly scanned point cloud data to the total point cloud model, resulting in too large data volume, or data redundancy in local areas

Images