GEOSECTIONING DNC Data

1
GEOSECTIONING DNC Data

• Prototype and Development Project
• Phase V

David Turnbull 03 May 2006 NGA HydroVision
Production Cell
2
Goals of Effort

• Have an in-house automated capability to convert
  DNC data into a one-feature-one-time format to
  easily incorporate into any database.
• Create and demonstrate a new prototype version of
  the DNC where it is one-feature-one-time.
• Gain the many side benefits created by having
  these automated tools.

3
Concepts

• Geo-sectioning
• One Feature One Time
• New Attribution to DNC
• Generalization and SCAMIN
• Common tile sizes
• Common Schema
• Libraries based on one or more 1 degree tiles

4
SCAMIN (Scale Minimum)

• SCAMIN is used in some degree with the
  international community with S-57.
• SCAMIN is the least scale in which a feature
  should be portrayed
• Currently based mostly on HACG scale ranges
• Harbor 0-49,999
• Approach 0-99,999
• Coastal group 1 0-249,999
• Coastal group 2 0-499,999
• General 0-2,500,000

5
Loaded 29 CDs into single Database

• Development with Laser-Scan, Inc. (LSI) which
  gave Lamps2 an automated mass importer of DNC
  CDs into a Gothic Object-Oriented (OO) Database
• Loaded all 29 CDs into individual databases
  within the database in the NGA HydroVision
  Production Cell (NPC).
• A single CD can be imported with a few clicks-
• average time for a CD 40 minutes
• NPC has the latest of all the CDs loaded into the
  Gothic database
• Also can run batch processing on multiple
  libraries

6
Propagated Into Single Dataset

• Chose the entire DNC17 since we had been
  maintaining it (last 3 phases as well)
• All DNC17 data brought into a single dataset
  using a fully automated tool created within
  Lamps2

7
New Attributes Added During Feature Propagation

• New attribution is per individual feature or
  object and added automatically
• New attributes
• Library Name-Name of the original library
• DNC17_h1708210_Baltimore
• Chart Identifier-Chart number feature came from
• Chart Edition Number-Edition of Chart
• Chart Scale (NEW FOR THIS PHASE)
• Hydrographic Datum (soundings only)

8
Geo-sectioned and Generalized Data

• Originally manual
• Now fully automated process (one click)
• Step 1-initial SCAMIN attribute created and
  populated per feature based on library type.
• Step 2- Pub Aid Number and Pub Number attributes
  added and populated from Text Attribute for
  Buoys, Beacons, and Lights
• Step 3-Uses the Data Quality Areas (DQY) to
  spatially select the features and splits all line
  and area features at DQY boundaries.

9
Geo-sectioned the Data (2)

• Step 4- Process features by DQY and search within
  a defined radius to see if there are matching
  features of smaller scale. If there are, the
  SCAMIN attribute value is replaced with the
  smaller scale. (Point features only)
• Lights-Additionally compares Publication Aid
  Number (PAN) and Range. If the range is not equal
  it looks at the PAN. If the PANs are equal it is
  a matching feature. If the PANs are not equal it
  produces a new QR light feature so that the
  compiler can visit it and determine if it is a
  different light or an error.

10
Geo-sectioned the Data (3)

• Step 5-Deletes unwanted, smaller scale,
  overlapped features.
• Only the best scale features are kept within the
  DQY.
• The data is now one-feature-one-time in a single
  dataset

11
Created the New Tiles

• Created Tiles within the new dataset using
  automated tile creator developed in the NPC using
  LULL, all tiles 1x1.
• Areas which originally covered Harbors and
  Approach Datasets are now based on a 1 x 1 degree
  tile scheme.
• Coastal and General areas are a maximum of 3 x
  3 deg and a maximum of 9 1x1 deg tiles.
• Tiles do not overlap, libraries do not overlap
• Split data along new library boundaries (auto)

12
New Datasets Based on Tiles

• Each library will be made up of either 1 or many
  tiles up to a max of 9.
• DNC17 divided into 246 tiles, 60 datasets
• The new datasets are named incorporating the tile
  name, ex h17hjfk
• the h so that it will meet spec. All libraries
  start with h (easier for us to code for now)
• 17 for DNC17
• hjfk for the tile that makes up the library

13
New Libraries For DNC17
14
Propagated Data to Datasets

• Spatially queried each tile area and propagated
  the data from the single master dataset into each
  new tile library
• Now fully automated for this phase

15
Modified Files for Export

• Changed export scripts to effect all libraries
  to be exported
• Allows for export of the new attributes and tile
  sizes, now automated
• Easy to modify the code/schema to accomplish this.

16
Exported Data

• All 60 libraries exported without viewing errors.
• A cleanup process would be needed if we were to
  use it for production (edge-matching)
• New attributes appeared properly structured
  within the VPF tables.
• Created a Browse for the libraries (now
  automated)
• Cut a CD of the new Prototype IV for DNC17.

17
Test on ECDIS platforms

• Gave to SPAWAR to modify COGENT to view the data
  and utilize SCAMIN.
• Data views utilizing SCAMIN.
• Load into various ECDIS platforms
• FUND Testing
• ECPINS Testing
• ICE Testing
• VMS Testing

18
Demonstration of SCAMIN

• Created customized code within Lamps2 to
  demonstrate the SCAMIN value being used to thin
  out the visible data when zooming out/in.
• Proof of concept, the ECDIS would have to be
  modified to utilize the SCAMIN attribute as
  SPAWAR has.

19
Depth Distribution
20
500,000 view
21
250,000 view
22
100,000 view
23
50,000 view
24
25,000 view
25
10,000 view
26
Benefits

• We now can go to one-feature-one-time
  automatically in-house in order to populate the
  future NDME or NGA databases.
• Proved that adding new attributes to the data
  properly will not kill the ECDIS functionality.
• Size of the data on a CD is smaller.
• Normal DNC17 415 MB
• Geo-sectioned DNC17 274 MB
• No duplication in collecting.

27
Side Benefits

• Have used it already for the foundation for the
  EPODS effort to supply one-feature-one-time data
• Have used it to supply custom datasets to several
  customers (NAVO, TRANSCOM, HYSAS)
• Utilized tools to create a One-Feature-One-Time
  Global DNC coastline.
• In process of using tools to create an ArcGlobe
  global DNC model.

28
What is nextPhase V

• Further testing on ECDIS platforms
• Finish display methods in Lamps2 to utilize the
  SCAMIN values for lines and areas.
• Possibly Adding Chart limit feature class for
  Hardcopy.
• Possibly creating a model using more SCAMIN bands
  similar to the IHO models
• Possibly incorporating new agreed upon attributes
  from DNC2/Data Model/S57 efforts

29

• QUESTIONS?

30
NATIONAL GEOSPATIAL-INTELLIGENCE AGENCY