International Archives of Photogrammetry and Remote Sensing, Vol. 32, Part 3W14, La Jolla, CA, 9-11 Nov. 1999 
     
PHOTOGRAMMETRY AND LASER ALTIMETRY 
Toni Schenk 
Department of Civil and Environmental Engineering and Geodetic Science, OSU 
schenk.2@osu.edu 
KEY WORDS: Photogrammetry, Laser Ranging, DTM generation, Surface Reconstruction, Calibration, Segmentation, 
Fusion 
ABSTRACT 
Laser altimetry has, and is continuing to have a profound impact on photogrammetry. This paper discusses the 
interrelationship between the two fields, identifies problems, and suggests a common research agenda. Surface 
reconstruction is an important application in photogrammetry. DTMs, for example, are as much an end result as 
an intermediate step for orthophoto generation and for object recognition. The quality control is a central issue, 
regardless how the surface is obtained. The paper analyzes the effect of systematic errors of laser ranging systems 
on the reconstructed surface. To keep these effects as small as possible, rigorous calibration methods are necessary, 
including suitable test surfaces and proper error modeling. In a more complex setting, better results are achieved 
if laser ranging is combined with stereopsis. The combination of different surface measurements is a challenging 
fusion problem. At the sensor level, the systems, such as camera and laser, can be tightly integrated to the degree 
that the laser footprint is recorded by the camera. On the feature level, extracted features from imagery and laser 
ranging must be combined in a way that accounts for the different error models. Finally, on the symbolic level, a 
combined surface must explain the measurements resulting from the different sources. 
1 Introduction 
Photogrammetry and airborne laser ranging (ALR) are 
the two most widely used methods for generating dig- 
ital terrain (DTMs) and surface models (DSMs). Surfaces 
play an important role in photogrammetry. A DTM is 
as much an end result as an intermediate step for or- 
thophoto production and object recognition. 
The development of airborne laser ranging started in 
the 1970s in North America mainly for bathymetric ap- 
plications. With the emergence of GPS and INS other 
applications, such as monitoring ice sheets in Green- 
land and measuring canopy heights were successfully 
approached. As a result, a "laser altimetry" community 
emerged in North America with researchers from NASA 
forming the nucleus, and photogrammetrists being vir- 
tually absent—quite in contrast to Europe where pho- 
togrammetrists greatly facilitated the application of air- 
borne laser ranging. 
The present workshop Mapping Surface Structure and 
Topography by Airborne and Spaceborne Lasers is 
unique in that it brings together for the first time 
the laser ranging and photogrammetry community— 
undoubtedly, for the benefit of both. This paper dis- 
cusses the interrelationship between the two fields, 
identifies problems and suggests a common research 
agenda. 
The paper begins with a comparison of photogramme- 
try and laser ranging with respect to generating DTMs. 
Such comparisons have been carried out in the past, see, 
e.g. Ackermann (1999) and Baltsavias (1999) for an ex- 
tensive review. Here, we emphasize similarities and im- 
portant differences for the motivation of a joint research 
agenda. The comparison is divided into data acquisition 
and surface reconstruction. 
Some effort is placed on the impact of systematic er- 
rors on the reconstruction of surfaces from laser rang- 
ing. This is an important issue because laser ranging 
does not provide redundant information for the compu- 
tation of 3-D points. Whether a point is correct or not 
can only be inferred later—if at all. The quality of laser 
surfaces depends a great deal on a well calibrated sys- 
tem, including the identification of systematic errors. 
The raw laser data forms a cloud of 3-D points that have 
no topological relationship. Surface properties, such 
as breaklines and abrupt discontinuities in surface nor- 
mals, must be made explicit. This is an important post- 
processing task. The remainder of the paper is a brief 
summary of essential post-processing steps. Details 
about most aspects appear as separate papers from our 
research group at OSU. The collaborative work of Byrd 
Polar Research Center's Ice Dynamic Laboratory, headed 
by B. Csathó, and the digital photogrammetry labora- 
tory are a most rewarding experience. We hope that this 
example encourages further contacts between the laser 
and photogrammetry community in North America. 
2 Comparison between Photogrammetry and Laser 
Ranging 
Photogrammetry and airborne laser ranging (ALR) are 
the two most popular methods for generating digital 
terrain models (DTMs) or digital surface models (DSMs) 
of extended areas. Hence it stands to reason to com- 
pare the two methods. This section elucidates similar- 
ities and relevant differences. In the interest of brevity 
we keep the comparison on a general level here. Acker- 
mann (1999) and Baltsavias (1999) provide more de- 
tailed analysis, for example. Table 1 contains a number 
of comparison factors that are divided into data acquisi- 
tion and surface reconstruction.