Proceedings 18 th International Symposium CIPA 2001 
Potsdam (Germany), September 18 - 21, 2001 
SUPPLEMENTING LASERSCANNER GEOMETRIC DATA 
WITH PHOTOGRAMMETRIC IMAGES FOR MODELING 
Fredie Kern 
Institute for Geodesy and Photogrammetry, Technical University of Braunschweig, Germany 
GauBstraBe 22, D-38106 Braunschweig 
f.kern@tu-bs.de 
KEY WORDS: Laserscanner, Camera Calibration, Texture Mapping 
ABSTRACT 
The laser scanner take just a few minutes time to scan millions of points with arbitrary resolution. These can be compared to results 
achieved from measurements done using photogrammetric standard. Although photogrammetric methods produce both geometric 
and color information while scanner can up to now not capture color. Furthermore, one can determine easily interpret and analyse 
photos thus being able to extract further details like the object material or supplementing the geometry of areas in the shadow of the 
objects. The paper should explain how laserscanning can be used to fill up the photometric shortfalls in the use of two pictures for 3D 
geometric measurements. The combination of photogrammetric and laser scanner systems compliments one another perfectly for a 
complete object modelling. The main function of the combined system is to create textures for the surface model derived from a 
cloud of points. Therefore it is necessary to examine the mathematical function for the geometric relation between the images and the 
cloud of points. A calibration method for this problem will be described. An example for an automated generated 3D model is given. 
1. OBJECT MODELING 
Measurements are necessary to create a model of the reality. Models are used in documenting a fixed condition of the reality. By 
analyzing the model one can get some new knowledge about the object. Two different forms of information can be separated in 
modelling processes. There is on one hand, geometric information which is very important for object description, the geometry then 
forms the foundation of documenting further thematic information on the another hand. The thematic information could be data of 
physical, chemical, historic or legal properties. There exists a lot of technical solutions for the determination of these physical 
parameters, eg. kinds or strength of material. Photogrammetry is a well known technique for measurement of geometric information 
and physical properties which are influenced by light. With photos you can determine the object color perfectly. Despite all modern 
technology, a human expert has to finally survey all facts to examine the important object features like building epoch or construction 
details. 
2. PHOTOGRAMMETRY AND LASERSCANNING 
2.1 The Essence of Photogrammetry 
When we confine object modeling only to the geometric modeling, photogrammetry is an elegant measurement method. One is able 
to remotely capture a lot of images by occupying an object location for a short time period. Each image is a model of an object part in 
the form of an array of pixels, with the pixel values being a function of the object color and the lighting exposure of the object 
surface at the moment of photography. The image is a miniature model of the reality based on optical laws. In architectural 
photogrammetry, image scales range from 1:50 to 1:200 which implies that one pixel represents many object details. The image scale 
and the geometric pixel size define the object resolution. A big advantage of the photogrammetric method is that the central 
perspective is a simulation of the human seeing process thus making image interpretation by a viewer easy. On the contrary, it 
requires in theory two images to extract 3D information whereby relative strong restrictions must be fulfilled by the two images. 
They must be taken from different positions, cover same object parts and the rays to homologous points must intersect at a good 
angle. Usually more than two, typically more than a dozen images are used to reconstruct an object in 3D co-ordinates. Subsequent 
from the point information higher geometric primitives like lines or planes are modelled. This point by point evaluation results in a 
wireframe model or an boundary representation which can used to create ground plans, cuttings or views. 
2.2 The Essence of Lasersanning 
Laserscanners work in a similar manner like robotic tachymeter with reflectorless distance measurement capability (Fig. 1). A laser 
beam is sent out in all directions within a equidistant grid. The laser light is reflected back to the laserscanner by the object surface. 
By measurement of the time of flight the distance in a fixed direction can be determined. Two fast rotating mirror deflects the laser 
light in vertical and horizontal directions. Some laserscanner systems however deflect only in horizontal or vertical directions and 
realize the second component of direction (vertical resp. horizontal) with the help of a servo. The deflection mirror and servo system 
in a laserscanner correspond to the readings (horizontal and vertical angles) of the pitch circles in a theodolite. With the distance 
measurement and the corresponding horizontal and vertical angle the 3D position of the object can be derived. The angle of the 
laserscanner beam radiation defines the resolution of the scan. The scanning is an automatic process where the whole sphere around 
the scanner is measured resulting in a cloud of points.