Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

Digital map database quality evaluation issues for ITS applications

Aggarwala, R. K., Nystuen, J. D., Frank, Andrea Irmgard, Palathinkara, J. and Frank, L. 1997. Digital map database quality evaluation issues for ITS applications. Presented at: UCGIS Annual Assembly and Summer Retreat, Bar Harbor, MA, USA, 21-24 June 1997.

Full text not available from this repository.

Abstract

The paper reports ongoing research in which we are developing tests and procedures to evaluate quality of digital map databases for Intelligent Transportation Systems (ITS) applications. One of our frames of reference is the Society of Automotive Engineer's standard entitled, Truth-in-Labeling Standard for Navigation Map Databases (SAE J1663, issued August 1995). Broader issues of database standards, interoperability across regions, and between ITS products are also involved. Digital map databases are representations of entities in the real world in which locational references, attributes and relationships are preserved. The representations must meet accuracy and functional requirements which, in turn, depend upon the type of applications. We are developing methods for identifying and separating different types of errors on a vendor map by comparing segments of the vendor map with the corresponding portion of a reference map of greater positional accuracy. So far our work has focused on positional and topological measures of errors in a digital map intended for use for motor vehicle navigation. Errors of displacement, omission and commission are identified by these procedures. We have found that when discrepancies of definition/representation and currency exist with respect to the two sources, the discrepancies cannot be distinguished from errors due to interdependencies, hence the pressing need for data standards. We have tried to automate the error testing procedures leaving human interpretation and ground verification to a minimum in order to keep costs of quality checking at some feasible level. We have found that the positional errors are not just randomly distributed but have associated systematic errors, and possess high spatial variability. Further study of the spatial structure of this distribution is required in order to develop an appropriate sampling strategy.

Item Type: Conference or Workshop Item (Paper)
Date Type: Publication
Status: Unpublished
Schools: Geography and Planning (GEOPL)
Subjects: G Geography. Anthropology. Recreation > G Geography (General)
Q Science > QA Mathematics > QA76 Computer software
Last Modified: 07 Nov 2019 09:10
URI: http://orca.cf.ac.uk/id/eprint/42110

Actions (repository staff only)

Edit Item Edit Item