KEYWORDS: Clouds, Data modeling, 3D modeling, Reconstruction algorithms, Raster graphics, 3D scanning, Local area networks, Laser scanners, Laser applications, Image segmentation
An iterative slicing reconstruction method for point cloud surface holes is proposed to address the problem that the traditional hole repair method fails in repairing surface holes with uneven density. Firstly, the least squares micro-slices are used to detect and extract the point cloud hole boundaries, and then the least enclosing box is constructed and initially rasterized to achieve a uniform segmentation effect. Then the density of segmentation results is analyzed and judged, and if the density is too large, iterative slicing calculation is performed to obtain uniformly dense segmentation blocks. Finally, the moving least squares method is used to fit each slice data to reconstruct the missing part of the point cloud surface. Our results show that this method can achieve the effect of filling the point cloud holes and averaging the point cloud density as well as improving the accuracy of hole repair for holes containing curved surfaces or point cloud data with uneven density.
KEYWORDS: Clouds, Image segmentation, Reconstruction algorithms, Data modeling, 3D scanning, 3D modeling, Nonlinear filtering, Feature extraction, Local area networks, Image processing
Accurate segmentation of building facade point clouds is the key to 3D reconstruction of buildings. The region growing algorithm is widely used because of its simplicity and ease, but the traditional region growing algorithm leads to over-segmentation and under-segmentation problems due to the low robustness of seed points and the large differences in local features of building facade point clouds. To address the above problems, this paper proposes a building facade division based on FPFH feature classification and regional growth. Firstly, Kd-Tree is constructed to spatially index the facade point clouds and construct geometric topological relationships for the cluttered point clouds. Then, FPFH feature values are calculated and sorted for classification, while the points with relatively low feature values are selected as the initial seed points to ensure the stability of the seed points. Finally, the initial seed points are used as the reference for regional growth and face slice segmentation of the building facade point cloud. The experimental results show that the correct rate of the method in this paper is improved by 14.60%, the over-segmentation rate is reduced by 86.20%andtheunder-segmentation rate is reduced by 43.13% compared with the region growing algorithm, which not only improves the over-segmentation and under-segmentation problems, but also increases the segmentation accuracy as well as efficiency.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.