Share Email Print

Proceedings Paper

Merging vector features extracted from imagery with GIS data using vector linking
Author(s): Robine Lewis
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Development of vector feature extraction techniques combined with large amounts of digital map data can create a rich database for intelligence activities. Our ability to use this data depends on understanding the relationships between different features and different representations of the same feature. Linking is the process of determining that two features in different layers represent the same object. Links between features in two or more layers can be used for: co-registering vector data from multiple sources, automatic revisions of linear data, feature attribute inheritance, or network transversal problems. One sample application uses linking for generating a database with features extracted from stereo imagery and attributes from existing DMA data sources. Time and personnel are limited resources. Therefore, the linking process needs to be automated. Two data sets are used -- pre-existing road data at 1:250 K scale (Set A) and stereo imagery used for extracting 1:50 K roads (Set B). Set A is richly attributed. We want the spatial accuracy of the roads from Set B and the attributes of the roads from Set A. Linking can match the two sets of roads. A new combined data set is created in several stages. Roads where a one-to-one linkage exists between A and B use the spatial data from Set B and the attributes from Set A. Roads that are unique in either set are added. Attributes are retained for the roads from Set A. This paper discusses a new technique for automatic feature linking developed at GDE Systems Inc. and demonstrated in a prototype. The prototype uses the characteristics of a linear vector feature to identify the same features from different sources and from different scales.

Paper Details

Date Published: 31 December 1996
PDF: 8 pages
Proc. SPIE 2960, Remote Sensing for Geography, Geology, Land Planning, and Cultural Heritage, (31 December 1996); doi: 10.1117/12.262469
Show Author Affiliations
Robine Lewis, GDE Systems, Inc. (United States)

Published in SPIE Proceedings Vol. 2960:
Remote Sensing for Geography, Geology, Land Planning, and Cultural Heritage
Daniel Arroyo-Bishop; Roberto Carla; Joan B. Lurie; Carlo M. Marino; A. Panunzi; James J. Pearson; Eugenio Zilioli, Editor(s)

© SPIE. Terms of Use
Back to Top