Knowledge-based automated road network extraction system using multispectral images

Weihua Sun; David W. Messinger

doi:10.1117/1.OE.52.4.047203

15 April 2013 Knowledge-based automated road network extraction system using multispectral images

Weihua Sun, David W. Messinger

Author Affiliations +

Optical Engineering, Vol. 52, Issue 4, 047203 (April 2013). https://doi.org/10.1117/1.OE.52.4.047203

Abstract

A novel approach for automated road network extraction from multispectral WorldView-2 imagery using a knowledge-based system is presented. This approach uses a multispectral flood-fill technique to extract asphalt pixels from satellite images; it follows by identifying prominent curvilinear structures using template matching. The extracted curvilinear structures provide an initial estimate of the road network, which is refined by the knowledge-based system. This system breaks the curvilinear structures into small segments and then groups them using a set of well-defined rules; a saliency check is then performed to prune the road segments. As a final step, these segments, carrying road width and orientation information, can be reconstructed to generate a proper road map. The approach is shown to perform well with various urban and suburban scenes. It can also be deployed to extract the road network in large-scale scenes.

Citation Download Citation

Weihua Sun and David W. Messinger "Knowledge-based automated road network extraction system using multispectral images," Optical Engineering 52(4), 047203 (15 April 2013). https://doi.org/10.1117/1.OE.52.4.047203

Published: 15 April 2013

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available