URBAN ORTHOIMAGE ANALYSIS GENERATED FROM IKONOS DATA 
S. Siachalou 
Dept. Cadastre, Photogrammetry and Cartography, Aristotle University of Thessaloniki, GREECE, sofiasiahalou@hotmail.com 
Youth Forum 
KEY WORDS: Remote Sensing, Ikonos, orthoimage, fusion, urban, 3D visualization. 
ABSTRACT: 
Due to the increasing demand and use of digital spatial information, the interest is being drawn to the generation of digital 
orthophotos, which is the most accurate and reliable source of spatial information. The subject of this project is the generation of 
orthoimages as well as the further processing of these orthoimages for the full exploitation of their spectral and spatial information. 
The imagery data used are a panchromatic image of one-meter spatial resolution and a multispectral image of four-meter spatial 
resolution, both acquired from Ikonos-2 satellite over the extended region of the city of Thessaloniki. The orthorectification of the 
images involved the establishment of the interior and exterior orientation through the Rational Function coefficients, a number of 
ground control points, and a Digital Terrain Model (DTM). Following the orthorectification, the panchromatic orthoimage was fused 
with the multispectral orthoimage to produce a pan-sharpened image with the method Principal Components Transformation. At this 
point the spectral quality and the spatial accuracy of the pan-sharpened image were assessed by certain criteria, as it is very 
significant for the sharpened image to maintain the spectral and the spatial information of the original data. In order to produce an 
image containing thematic information a supervised classification pan-sharpened was realised. Finally the products of the analysis of 
the pan-sharpened orthoimage were used for the visualization of the study area. 
1. INTRODUCTION obtained by GPS measurements and orthophotos of scale 1: 
5000. 
Since the launch of Ikonos satellite a new era for Remote 
Sensing products has began. The generation of high-resolution The software used for the processing of the data is Erdas 
orthoimages is an important task as it has a use in various IMAGINE 8.5. 
applications such as mapping, agriculture and urban planning. 
Thus it is of great importance to produce digital spatial products 
and improve their geopositional accuracy. In this project the 
improvement of the accuracy of the orthoimages was 
accomplished by the use of GCPs obtained by GPS 
measurements and orthophotos, a DTM and Ikonos images. 
Since the camera model and precise satellite ephemeris data are 
not available for Ikonos imagery, the use of the Rational 
Function was decided as it can offer a very accurate 
approximation to the rigorous physical sensor model (Kratky, 
1989). 
  
  
  
  
  
Furthermore the applications mentioned require further 
processing of the orthoimages in order to extract all the useful 
information. Having this in mind, the panchromatic orthoimage 
was fused with the multispectral orthoimage so as to achieve 
high spatial resolution while maintaining the provided spectral Figure 1. a) Map of Greece with the area of interest b) the 
resolution,. It should be noted that the synthetic orthoimage is panchromatic lkonos image 
ideal for the extraction of thematic information, which can be 
made though classification. 
  
  
  
  
  
3. ORTHORECTIFICATION 
2. DATA 3.1 Geometric Sensor Models 
Sensor models are vital in orthorectification because they 
describe the relationship between the coordinates of the image 
space and the object space and allow their transformation. 
There are two categories of sensor models, the physical and the 
generalised models. The physical sensor model establishes the 
physical imaging process. This can be a disadvantage because 
the position and orientation of a satellite sensor changes during 
the image acquisition, the geometric model is time-dependent 
The satellite data used is a panchromatic Ikonos image of 1-m 
resolution and a multispectral Ikonos image of 4-m resolution 
both over the extended region of the city of Thessaloniki. For 
the orthorectification of these images two DTMs were 
available, the first one has a grid size of 25-m and covers all the 
study area while the second one has a grid size of 20- m and 
covers the hilly region of Seih Sou. The GCPs used were 
186