eprintid: 37133
rev_number: 12
eprint_status: archive
userid: 5878
dir: disk0/00/03/71/33
datestamp: 2025-08-12 08:09:57
lastmod: 2025-08-12 08:10:25
status_changed: 2025-08-12 08:09:57
type: article
metadata_visibility: show
creators_name: Tomljenović, Ivan
creators_name: Blaschke, Thomas
creators_name: Höfle, Bernhard
creators_name: Tiede, Dirk
title: Potential and idiosyncrasy of object‐based image analysis for airborne Lidar‐based building detection
subjects: ddc-550
divisions: i-120700
keywords: Remote Sensing, Geography, LiDAR, Image processing and analysis
abstract: Today, a plethora of approaches exists that can be used for the task of object 
extraction from LiDAR data (point and/or derived raster information). The majority of 
approaches deal with the extraction of artificial objects or distinct natural objects (e.g. single 
trees). Local height variations or number of signal returns are often input to extraction 
processes. If only relying on pixel‐based concepts, results of such analyses may provide 
sufficient detection accuracy but they may also generate various misclassified objects since 
they lack object‐based view. In the presented paper, we developed an Object‐Based Image 
Analysis (OBIA) approach for building extraction starting from LiDAR point data only. We 
hypothesize that an OBIA approach could performs better than a pixel‐based approach as it 
resembles the way in which humans perceive reality. The developed method consists of a 
set of rules applied to continuously re‐shaped (based on pre‐classifications) candidate 
objects. The results of the proposed approach have been compared to buildings which were 
manually identified from a Digital Surface Model (DSM). Primary results show completeness 
value of 97.80% and correctness of 80.5%.
date: 2014
publisher: Aristotle University of Thessaloniki
id_scheme: DOI
id_number: 10.11588/heidok.00037133
own_urn: urn:nbn:de:bsz:16-heidok-371333
language: eng
bibsort: TOMLJENOVIPOTENTIALA2014
full_text_status: public
publication: South-Eastern European journal of earth observation and geomatics
volume: 3
number: 2S
place_of_pub: Thessaloniki
pagerange: 517-520
issn: 2241-1224
citation:   Tomljenović, Ivan ; Blaschke, Thomas ; Höfle, Bernhard ; Tiede, Dirk  (2014) Potential and idiosyncrasy of object‐based image analysis for airborne Lidar‐based building detection.  South-Eastern European journal of earth observation and geomatics, 3 (2S).  pp. 517-520.  ISSN 2241-1224     
document_url: https://archiv.ub.uni-heidelberg.de/volltextserver/37133/1/tomljenovic_et_al_2014_geobia.pdf