Scribd
Upload
176 views42 pages

3 DTiles in Action

Sean Lilley presented on 3D Tiles, an open source format for streaming and rendering massive heterogeneous 3D geospatial datasets. 3D Tiles addresses issues with existing 3D data formats by defining a tiling scheme with level-of-detail and binary tile formats like batched, instanced, point cloud and vector. This allows for fast loading and rendering of large datasets across different 3D visualization engines. Current work is focused on a truly 3D vector tile format, an analytics-enabled styling language, and time-dynamic streaming capabilities.

Uploaded by

Eduardo Nuno
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
176 views42 pages

3 DTiles in Action

Sean Lilley presented on 3D Tiles, an open source format for streaming and rendering massive heterogeneous 3D geospatial datasets. 3D Tiles addresses issues with existing 3D data formats by defining a tiling scheme with level-of-detail and binary tile formats like batched, instanced, point cloud and vector. This allows for fast loading and rendering of large datasets across different 3D visualization engines. Current work is focused on a truly 3D vector tile format, an analytics-enabled styling language, and time-dynamic streaming capabilities.

Uploaded by

Eduardo Nuno
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 42

IN ACTION

Sean Lilley
Analytical Graphics, Inc
[email protected]
@lilleyse
• Open source JavaScript library for 3D globes & maps
• Built on WebGL
• Founded by Analytical Graphics Inc.
3D Data
• City building data
• Photogrammetry
• LiDAR
• Vector data
• CAD & BIM
3D Data
• City building data – CityGML, GeoJSON, Shapefile, KML + COLLADA
• Photogrammetry – COLLADA, OBJ, hundreds of jpegs
• LiDAR – LAS, PCD, CSV with xyz/rgb
• Vector data – GeoJSON, Shapefile, KML
• CAD & BIM – IFC, DWG, other proprietary formats
3D Data
• Many input formats
• Common for datasets to be in the multi GB range
• Inconsistent projections
• Not inherently streamable
• Amount of data is only growing
3D Data Case Studies
Melbourne
• One hundred 8k by 8k images
• 100 million triangles
• 136 obj models
Cell Tower
• 900 images
• 5.6 million triangles
Tiling in 3D is a necessity
Vulci Site
• 9634 images
NYC
• 1.1 million buildings
• 12 GB CityGML
swisstopo
• 40 million buildings
• 80 million trees
• 150,000 labels
• KML + COLLADA
Desert Point Cloud
• 1.5 GB LAS file
• 55 million points
Philly Trees
• 57K trees
• Instanced from single
COLLADA model
• CSV file with lat/long
3D Data
• Too many input formats for renderers to realistically support
• Not optimized for streaming or rendering
• What we need:
• A spatial data structure with level-of-detail
• Geometry and texture simplification for non-leaf tiles
• Quick to offload to GPU
Massive heterogeneous 3D engines, e.g.,
3D geospatial datasets
3D Tiles
• tileset.json
• Defines spatial layout for the tileset
• Relative paths to tiles
• May point to external tilesets
• Tiles
• Contains the actual data
• Binary formats

• Alternatively – self contained .3dtiles SQLite DB


Tile Formats
• Batched (b3dm)
• Instanced (i3dm)
• Point Cloud (pnts)
• Vector (vctr)
Tile Formats
• Batched (b3dm)
• Instanced (i3dm)
• Point Cloud (pnts)
• Vector (vctr)
Tile Formats
• Batched (b3dm)
• Instanced (i3dm)
• Point Cloud (pnts)
• Vector (vctr)
Tile Formats
• Batched (b3dm)
• Instanced (i3dm)
• Point Cloud (pnts)
• Vector (vctr)
Tile Formats
• Batched (b3dm)
• Instanced (i3dm)
• Point Cloud (pnts)
• Vector (vctr)
3D Data
• City building data – CityGML, GeoJSON, Shapefile, KML + COLLADA
• Photogrammetry – COLLADA, OBJ, hundreds of jpegs
• LiDAR – LAS, PCD, CSV with xyz/rgb
• Vector data – GeoJSON, Shapefile, KML
• CAD & BIM – IFC, DWG, other proprietary formats
3D Tiles
• City building data – b3dm
• Photogrammetry – b3dm
• LiDAR – pnts
• Vector data – vctr
• CAD & BIM – b3dm
3D Tiles
• City building data – b3dm
• Photogrammetry – b3dm
• LiDAR – pnts
• Vector data – vctr
• CAD & BIM – b3dm

• Instanced features – i3dm


3D Tiles
• Different tile formats may exist in the same tileset
• Buildings + trees + point clouds
• Unified styling API
• Metadata layout is the same for each format
• In the end, everything is a feature
Vector Tiles + Buildings
Vector Tiles + Photogrammetry

Debug View
Style by height
Style by latitude
Style by distance to landmark
Hide small buildings
Point Cloud Styling
Since last year…
• Revamped styling API
• Batched, Instanced, and Point Cloud format in solid shape for 1.0
• Progress on vector format and classification
• Cesium implementation merged into master
• 3D Tiles in the OGC Community Standard Process
3D Tiles Ecosystem
Exporters / Tilers Visualization engines

three.js
prototype

OSG.JS

considering
ContextCapture MicroStation

FME

Validator in-progress

@mattshax

Built on
Next
A truly 3D vector tile for heterogeneous classification
A 3D analytics-enabled styling language
Time-dynamic streaming
3D Data
• City building data
• Photogrammetry
• LiDAR
• Vector data
• CAD & BIM
Thank you!

• https://github.com/AnalyticalGraphicsInc/3d-tiles
• Sean Lilley
• Analytical Graphics, Inc
• [email protected]
• @lilleyse

You might also like