In: Wagner W., Szfltely, B. (eds.): ISPRS TC VII Symposium - 100 Years ISPRS, Vienna, Austria, July 5-7, 2010, IAPRS, Vol. XXXVIII, Part 7B 
299 
AN INVENTED APPROACH IN IMAGE REGISTRATION 
“NEW ERA IN PHOTOGRAMMETRY” 
A. S. Homainejad 
Independent Research Group on Geospatial, Tehran, I.R. IRAN - s_homain@yahoo.com 
KEY WORDS: Digital Image, Laser Scanning, Transformation, Ortho-image 
ABSTRACT: 
This paper will discuss an invented approach which focuses on providing an output from registration of an aerial 
photograph on a 3D model such as DTM, DMS, or 3D model. The approach has been developed in order to omit 
distortions from output and to increase the reliability. In contrast of other image registration methods which their 
outputs are an image, this approach provide a 3D model which can be used for mapping, visualising, 3D GIS, ortho 
rectify image. The approach is able to register a mono image on DEM, DTM, DMS, or 3D model with the minimum 
requirement to camera calibration parameters. The paper will give a discussion on analysis of registration of an aerial 
image on laser scanning data and registration of a terrestrial image on a 3D model by using this approach. 
1. INTRODUCTION 
Photogrammetry since its emergence has had at least two 
major transitions. The first transition was begun at late 
60s or early 70s when analytical stereoplotters were come 
into the market and edged out analogue stereoplotters. 
The second transition was started at mid 90s or late 90s 
when digital photogrammetry was emerged and gradually 
pushed analytical stereoplotters out. In digital 
photogrammetry system a simulated stereoplotter 
digitally created in the computer and all processing have 
been implemented digitally, but the mathematical 
concepts are the same. Consequently, all processes in 
digital photogrammetry were computerised and a 
supplementary image processing package was integrated 
to the system. However, before emergence of digital 
photogrammetry, digital images, digital image 
processing, and digital image transformation were 
successfully implemented in remote sensing and in digital 
close range photogrammetry as well. 
For ascertaining a reliable photogramme trie output, 
enormous studies on photos, cameras, and mathematical 
modelling have been fulfilled. With a glance research on 
all studies, it will be recognised that all those studies, 
despite of in which period they have been fulfilled, have 
focused on the determination of interior parameters. For 
example, Hallert (1968) and Hakkarainen and 
Rosenbruch (1982) measured angles of rays with help of 
a theodolite and collimator for defining the focal length. 
Brown (1956) developed analytical camera calibration, 
and Brown (1966) mathematically modelled the 
decentring distortion. Kenefick et al (1972) developed an 
analytical camera self-calibration with eight physical 
parameters. Fraser (1997) explained parameters and 
mathematical model for digital camera self-calibration. 
Amiri Parian and Gruen (2010) developed a method of 
self-calibration for panoramic camera. 
Photogrammetry has embraced of a number of physics’ 
laws and it needs to fully familiar with them for 
calibration. Investigation on these physical rules is not the 
scope of this paper. More details regarding to above 
issues can be found in photogrammetry and remote 
sensing textbook, for example Slama (1980) and 
Lillesand and Kiefer (1987). Photogrammetry during its 
evolution has tested and implemented different 
approaches for coping with these issues and producing a 
reliable and precise output. 
This paper will give a report on implementation of the 
developed approach in two different projects. The first 
project focused on registering an aerial image on a laser 
scanning data and the second project registered a 
terrestrial image on a 3D model. 
This paper has been organised as follows. A brief 
investigation on data fusion and the proposal of data 
fusion for this project will be given in chapter 2. Chapter 
3 will explain the study area. Chapter 4 will describe and 
evaluate tests and outputs. A conclusion and remarks will 
be given in chapter 5.