|H2020 EINFRA nD-PointCloud Proposal
|Deeply Integrated and Semantic Extended nD-PointCloud
We propose the nD-PointCloud model as a breakthrough in handling massive multidimensional point cloud data sets
in the whole range from data ingestion, data management, to data analytics and visualisation. The data represent
space, time and added information such as colour, material properties, velocity, etc. State of the art
spatio-temporal representations are based on either gridded (raster, voxel) or object (vector) models. In many
cases these representations are organized in a fixed number of levels of importance (detail/scale), which
introduces serious limitations: fixed level choices and data density jumps between levels. The proposed
nD-PointCloud model facilitates continuous instead of fixed levels of importance. The continuous level of
importance value of a point can be regarded as an added dimension to space and time.
Original and novel aspects of our nD-PointCloud model:
- defining organizing and property dimensions
- offering attribute granularity: from individual point to group level
- obtaining the continuous level of importance value for a point
- aggregating points to higher level points
- representing and assessing effect of different coordinate reference systems: spherical or Cartesian
We will explore: nD space filling curves, tree structures, etc. to realize the deep integration of space, time and
importance as basis for data organization and apply high performance/throughput computing for big data (trillions
of points). By enabling operations directly on the raw point cloud data, nD-PointCloud largely avoids and/or
alleviates the extract, transform, load hurdle, which is an increasingly serious problem in the era of big data.
The deep integration enables semantically richer point cloud data functionality, which broadens the opportunities
for knowledge discovery. We expect major advances in domains requiring lossless spatio-temporal data of extremely
high accuracy, such as geo information and astronomy which are used as Proof-of-Principle.