719 
ADVANTAGES AND DISADVANTAGES OF THE HOUGH TRANSFORMATION 
IN THE FRAME OF AUTOMATED BUILDING EXTRACTION 
G. Vozikis 3 ’*, J.Jansa b 
d GEOMET Ltd., Faneromenis 4 & Agamemnonos 11, GR - 15561 Holargos, GREECE - george.vozikis@geomet.gr 
b Institute of Photogrammetry and Remote Sensing, Vienna University of Technology, A-1040 Vienna, AUSTRIA - 
jj@ipf.tuwin.ac.at 
KEY WORDS: Photogrammetry, Remote Sensing, Reconstruction, Automation, Building, Edge, High Resolution 
ABSTRACT: 
• 
This paper investigates the outcome of building extraction techniques from optical aerial and spacebome high resolution data when 
applying Hough Transformation procedures. The proposed workflow consists of 4 major steps, namely the pre-processing of the data, 
nDSM (normalized Digital Surface Model) creation, building localization and building extraction. The advantages and disadvantages 
of this approach are discussed. The general conclusion is that the Hough Transformation is a very powerful tool for increasing the 
degree of automation during building extraction, while it is very robust against noisy data. Additionally the level of detail of the 
extracted buildings can easily be adjusted, but dark shadows in the images can make the algorithm produce erroneous results. 
1. INTRODUCTION 
Automated building extraction from high resolution image data 
(either airborne or spacebome) is becoming more and more 
mature. Each day new techniques are investigated and the 
results are getting more and more reliable, while the degree of 
automation increases. Each of the building extraction methods 
is of course connected to certain pros and cons. The use of the 
Hough Transformation has proven to be very promising tool in 
the frame of the automated creation of Digital City Models, by 
extracting building properties from optical data. 
This paper investigates the Hough-Transformation-approach in 
detail and points out its advantages and disadvantages. 
Furthermore, it concludes for which kind of data sets and 
accuracy pretensions this approach is recommendable. 
Moreover, the reachable degree of automation is also examined, 
in order to see how reliable results are that were produced 
without human interaction. 
Altogether, five different datasets, coming from airborne and 
spacebome sensors, were examined. These datasets depict 
urban regions with varying building densities. 
2. WORKFLOW 
The proposed workflow for automated building extraction from 
image data by employing the Hough Transformation has been 
thoroughly described in Vozikis (2004). Figure 1 shows the 
major steps of the process. 
All steps in this workflow are highly automated and human 
interaction is reduced to a minimum. 
satellite imagery the orientation approach is based on the RFM 
(Rational Function Model) (Vozikis et al., 2003). When dealing 
with aerial imagery it is made use of GPS/INS information in 
order perform direct georeferencing, and thus automated image 
triangulation (Scholten and Gwinner, 2003). The DSM 
extraction is performed by automated correlation procedures, 
which nowadays are very mature and produce very good results. 
In the following the 4 major steps are briefly described: 
2.1 Pre-Processing 
This step comprises the procedures from orientation of the input 
data up to the DSM (Digital Surface Model) creation. For VHR 
Figure 1: Proposed workflow for automated building extraction 
from image data 
2.2 nDSM Creation 
The goal is to derive the DTM (Digital Terrain Model) from the 
DSM and subtract it from the DSM in order to produce the so-