Abstract
Updated road databases are required for many purposes such as urban planning, disaster management, car navigation, route planning, traffic management and emergency handling. In the last decade, the improvement in spatial resolution of VHR civilian satellite sensors – as the main source of large scale mapping applications – was so considerable that GSD has become finer than size of common urban objects of interest such as building, trees and road parts. This technological advancement pushed the development of “Object-based Image Analysis (OBIA)” as an alternative to pixel-based image analysis methods. <br><br> Segmentation as one of the main stages of OBIA provides the image objects on which most of the following processes will be applied. Therefore, the success of an OBIA approach is strongly affected by the segmentation quality. In this paper, we propose a purpose-dependent refinement strategy in order to group road segments in urban areas using maximal similarity based region merging. For investigations with the proposed method, we use high resolution images of some urban sites. The promising results suggest that the proposed approach is applicable in grouping of road segments in urban areas.
Highlights
Updated road databases are required for many purposes such as urban planning, disaster management, car navigation systems, route planning, traffic management, emergency handling, etc. (Grote, et al, 2012)
We have presented a strategy for automating the grouping of road segments in VHR images based on a purpose dependent grouping approach
The method is an extended version of the Maximal Similarity based Region Merging (MSRM) algorithm proposed by (Ning, et al, 2010) which is customized for road network segmentation
Summary
Updated road databases are required for many purposes such as urban planning, disaster management, car navigation systems, route planning, traffic management, emergency handling, etc. (Grote, et al, 2012). Intensive research has been conducted on automatic road extraction from VHR optical images (Mayer, et al, 2006), SAR images (Hedman, et al, 2004; Saati, et al, 2015), LiDAR data (Samadzadegan, et al, 2009) and on the integration of different data sources (Rahimi, et al, 2015) It is still one of the important and challenging subjects in urban remote sensing. The spatial resolution (GSD) of VHR satellite sensors has become finer than the size of common urban objects of interest such as buildings, trees and road parts. These objects are imaged in several pixels. In object based classification processes spectral, textural, structural information as well as context can be used
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
