Identifying Planes in Point-Clouds for Efficient Hybrid Rendering

In proceedings of The 13th Pacific Conference on Computer Graphics and Applications, Oct. 2005
Presented at The 13th Pacific Conference on Computer Graphics and Applications (Pacific Graphics 2005)
 

Abstract

We present a hybrid rendering technique for highfeature colored point clouds that achieves both, high performance and high quality. Planar subsets in the point cloud are identified to drastically reduce the number of vertices, thus saving transformation bandwidth at the cost of the much higher fill-rate. Moreover, when rendering the planes, the filtering is comparable to elaborate point-rendering methods but significantly faster since it is supported in hardware. This way we achieve at least a 5 times higher performance than simple point rendering and a 40 times higher than a splatting technique with comparable quality. The preprocessing performance is orders of magnitude faster than comparable high quality point cloud simplification techniques.

The plane detection is based on the random sample consensus (RANSAC) pproach, which easily finds multiple structures without using the expensive Hough transform. Additionally, we use an octree in order to identify planar representations at different scales and accuracies for level-of-detail selection during rendering. The octree has the additional advantage of limiting the number of planar structures, thereby making their detection faster and more robust. Furthermore, the spatial subdivision facilitates handling out-of-core point clouds, both in preprocessing and rendering.

Keywords: hybrid rendering, level-of-detail, out-of-core, plane detection, point cloud, point rendering

Images

Download Paper

Download Paper

Bibtex

@INPROCEEDINGS{wahl-2005-hybrid,
      author = {Wahl, Roland and Guthe, Michael and Klein, Reinhard},
       title = {Identifying Planes in Point-Clouds for Efficient Hybrid Rendering},
   booktitle = {The 13th Pacific Conference on Computer Graphics and Applications},
        year = {2005},
       month = oct,
    keywords = {hybrid rendering, level-of-detail, out-of-core, plane detection, point cloud, point rendering},
    abstract = {We present a hybrid rendering technique for highfeature colored point clouds that achieves both,
                high performance and high quality. Planar subsets in the point cloud are identified to drastically
                reduce the number of vertices, thus saving transformation bandwidth at the cost of the much higher
                fill-rate. Moreover, when rendering the planes, the filtering is comparable to elaborate
                point-rendering methods but significantly faster since it is supported in hardware. This way we
                achieve at least a 5 times higher performance than simple point rendering and a 40 times higher than
                a splatting technique with comparable quality. The preprocessing performance is orders of magnitude
                faster than comparable high quality point cloud simplification techniques.
                
                The plane detection is based on the random sample consensus (RANSAC) pproach, which easily finds
                multiple structures without using the expensive Hough transform. Additionally, we use an octree in
                order to identify planar representations at different scales and accuracies for level-of-detail
                selection during rendering. The octree has the additional advantage of limiting the number of planar
                structures, thereby making their detection faster and more robust. Furthermore, the spatial
                subdivision facilitates handling out-of-core point clouds, both in preprocessing and rendering.},
  conference = {The 13th Pacific Conference on Computer Graphics and Applications (Pacific Graphics 2005)}
}