[Pointdown] standardized point cloud streaming

[Peter van Oosterom]

All,

It is great to see the interest, on the pointdown mail list, in 
standardized streaming of point clouds. This message expresses some of 
our ideas on the topic as a result of the (Netherlands) eScience project 
on the management of massive point clouds. We have discussed possible 
standardization of point clouds several time with the project partners 
(Fugro, Oracle, Netherlands eScience Center, TU Delft and 
Rijkswaterstaat) and considered two closely related levels of 
standardization:
1.    Database SQL (Structure Query Language) extension for point 
clouds, and
2.    Web Point Cloud Services (WPCS) for progressive transfer of point 
clouds.

 From our internal discussions we concluded that (initial) emphasis 
should be on web services. The end-users, from industry, government, or 
research, want interoperability and be able to combine point cloud data 
from various sources with functionality in tools form other vendors/ 
developers, or even own developed specific purpose applications.  As the 
focus is on geographic applications, these point cloud standards should 
fit and be complimentary to existing standards from the Open Geospatial 
Consortium (OGC) and/or the International Organisation for 
Standardisation/Technical Committee 211 (ISO TC211). Quite a number of 
aspects have to be covered, resp. taken into consideration:
1.    actual point representation xyz (a lot, used SRS, various base 
data types: int, float, double, number, or perhaps even varchar..)
2.    attributes per point (0 or more; e.g. intensity I, color RGB or 
classification, or ...); note that this might conceptually correspond to 
a 4D, 5D or higher dimensional point),
3.    fast access (based on spatial coherence) -> some blocking scheme 
(in 2D, 3D, …)
4.    space efficient encoding/storage/transfer -> compression 
techniques (exploiting spatial cohesion)
5.    data pyramid (LoD, multi-scale/vario-scale, perspective) support, 
combined with streaming/ progressive transfer
6.    temporal aspect, options for time per point (costly) or block 
(less refined)
7.    selection accuracies (blocks, individual points on 2D, 3D or nD 
query range/ other selection geometries)
8.    operators, functionality for specifying some manipulation, 
deriving attributes, specifying specific subsets, etc.

Assume that the standardization idea gets enough momentum, we would be 
very interested in active participation in such an effort (and would be 
in favour of have ISO or OGC as environment).

On behalf on our project team, kind regards, Peter van Oosterom.

PS. We are approaching the issue from the data management perspective, 
more info on our activities on pointclouds.nl

-- 
Peter van Oosterom          P.J.M.vanOosterom at tudelft.nl
Section GIS technology      (room 00-west-520) Department OTB
Faculty of Architecture and the Built Environment, TU Delft
tel (+31) 15 2786950        Julianalaan 134, 2628 BL Delft, NL
fax (+31) 15 2784422        P.O. Box 5043, 2600 GA Delft, NL
http://geomatics.tudelft.nl MSc Geomatics
http://www.msc-gima.nl      MSc GIMA (Geo-Info Management&Appl)
http://www.gdmc.nl