3. Istanbul 2004 
ity control and 
In: Proceedings 
iver, Colorado, 
interferometric 
| learned on the 
s of the 4th 
onference and 
mote Sensing, 
003. STAR-3i 
): more lessons 
Proceedings of 
oplications and 
iarleston, South 
aphic line map 
erometric SAR. 
and Exhibition, 
d S. Crawford, 
gital elevation 
using airborne 
Conference, St. 
Using break line information in filtering process of a Digital Surface Model 
BADEA Dragos * , Acad. Dr. Ing. JACOBSEN Karsten ** 
*Technical University of Civil Engineering Bucharest Romania, 
Faculty of Geodesy — Laboratory of Photogrammetry and Remote Sensing 
badea_dragos(@yahoo.com, 
** Universitat Hannover 
Institut for Photogrammetry and GeoInformation 
jacobsen(@ipi.uni-hannover.de 
KEY WORDS: photogrammetry, DEM/DTM, LIDAR, Raster, Generation, Method, Three-dimensional. 
ABSTRACT: 
The most important factor that everyone cares is that of the similarity of digital representations with the real terrain surface. 
The problem occurs when you try to eliminate off-terrain points like manmade objects and trees. Those objects not belonging to the bare 
earth can be eliminated through different filter methods. In a flat area, an object can be detected by an algorithm which analyses the height 
of the points in relation to the surrounding area. But, what is happening vhen the terrain surface is not flat? The same algorithm is not 
anymore suited to filter off-terrain objects from the area with the same threshold. You can have the surprise that this filter will eliminate 
points from bare earth surface. And you have to find a different approach to the problem. 
This paper-work demonstrate the benefits brought by introduction of break lines in filtering the Digital Surface Model to achieve a more 
accurate Digital Height Model. 
The methods of filtering and generation of DHM we use, were developed and implemented in software at Institute of Photogrammetry 
and Geolnformation at the University of Hannover (IPI). 
LIDAR 
In this study, Lidar is used for terrain data collection 
technique i.e. photogrammetry, for generation of Digital Height 
Model (DHM). 
A pulsed laser ranging system is mounted in an aircraft equipped 
with a precise kinematic GPS receiver and an Inertial Navigation 
System (INS). The signal is sent towards the earth where it is 
reflected by a feature back towards the aircraft. A receiver then 
captures the return pulse. Using accurate timing, the distance to the 
feature can be measured. By knowing a speed of the light and the 
time the signal takes to travel from the aircraft to the object and 
back to the aircraft, he distances can be computed. Using a 
rotating mirror inside the laser transmitter, the laser pulses can be 
made to sweep through an angle, tracing out a line on the ground. 
By reversing the direction of rotation at a selected angular interval, 
the laser pulses can be made to scan back and forth along a line. 
Errors in the location and orientation of the aircraft, the beam 
director angle, atmospheric refraction model and several other 
sources degrade the co-ordinates of the surface point to 5 to 10 
centimetres . An accuracy validation study showed that Lidar has 
the vertical accuracy of 10-20 centimetres and the horizontal 
accuracy of approximately 1 meter. 
Obtaining DHM from Lidar data 
Among their advantages, these systems afford the 
opportunity to collect terrain data about steep slopes and 
shadowed and inaccessible areas (such as large mud flats and tidal 
areas). Following the initial post processing and positional 
assurance, the Lidar data are filtered for noise removal and 
prepared as a file of x,y,z points. 
  
Figure 1 Theoretical pulse emitted from a Lidar system 
The imagery can be used to add break lines to the Lidar data to 
reveal the terrain more accurately. 
DHM GENERATION FROM LIDAR DATA AND 
FILTERING 
A method for the generation of DHM from airborne 
Lidar data is presented which was developed at IPI. The method 
distinguishes itself by using the filter in the same time with 
interpolation. 
The original data obtained by ALS express the surface of ground 
objects, not only the ground surface but also trees and roofs of 
buildings. These data are called digital surface model (DSM). It is 
“necessary to distinguish these ground objects and to create a digital 
height model (DHM) that expresses the ground elevation by