VIRTUOZO DIGITAL PHOTOGRAMMETRY SYSTEM 
AND ITS THEORETICAL FOUNDATION AND KEY ALGORITHMS 
Prof. Jianging Zhang, Prof. Zuxun Zhang, 
Dr. Weiming Shen and Zhihong Wang 
Wuhan Technical University of Surveying and Mapping 
39 Loyu road, Wuhan, Hubei, 430070 P. R. China 
Tel: +86 27 788 1292 Fax: +86 27 787 3091 
Commission ll, Working Group lI/III 
XVIII Congress of ISPRS 
KEY WORDS: Digital, Photogrammetry, Systems, Performance, Theory, Algorithms. 
ABSTRACT: 
In the 1990's, photogrammetry strides into the digital age, having seemingly outgrown the analog and analytical generations. 
Digital photogrammetry processes digital images by specialised computer software ultimately producing the required digital 
maps and data. Such software, when bundled with prescribed hardware is termed a Digital Photogrammetric Workstation 
or DPW. The fully automated production of topographic maps with DPW's is still limited to the production of orthophoto 
maps. During the evolution of photogrammetry into the digital phase, a DPW (the Wuhan Digital Automatic Mapping System 
-WuDAMS ) was successfully developed at the Wuhan Technical University of Surveying and Mapping (WTUSM) in P. R. 
China. This DPW has entered the market with the trade name VirtuoZo. In this paper, the main functions and the theoretical 
foundation of VirtuoZo are introduced , then, based on the analysis of the weaknesses of traditional image matching 
algorithms, the probabilistic relaxation and Hopfield neural network techniques are discussed highlighting their application 
in image matching. Using these techniques, the spatial relationship and global consistency have been improved greatly. 
Based on the neural network method, zero matching can be contained, and break-lines and occlusions can be processed. 
The stability of image matching has been enhanced. For a practical DPW, effectiveness is an important factor. The 
algorithms in three of the most innovative modules of VirtuoZo are introduced. They can enhance the effectiveness of 
  
VirtuoZo by a factor five. Some of the applications of VirtuoZo are also outlined before a short conclusion. 
1. INTRODUCTION 
The development of photogrammetry has experienced three 
successive stages, namely the analog stage, analytical stage and 
digital stage. During the period of the analog stage (from about 
1900 to 1965), the typical photogrammetric instrument was the 
universal stereoplotter. This piece of equipment, simulated the 
bundles of straight lines through an object point, the perspective 
centre in the camera and its corresponding image points. This was 
achieved by optical, mechanical or optical-mechanical equipment. 
The analytical stage of photogrammetry started in 1957 when the 
concept of the analytical stereoplotter was proposed by U. V. 
Helava. Although the collinear equations were computed by a 
computer instead of analog projectors, the manual operation was 
nearly the same as in the universal stereoplotter, and the processed 
materials were still photographs. An advance was digital output in 
the form of the Digital Elevation Model ( DEM ) and mapping 
file, which was stored in the computer. This could be transmitted 
to a digital plotter to produce hardcopy output, as was required. 
This technique could be referred to as computer aided mapping. 
In the 1990's, photogrammetry strides into the digital age. Digital 
photogrammetry is referred to variously as pixel photogrammetry, 
softcopy photogrammetry, soft photogrammetry, rastergrammetry, 
fully digital photogrammetry or (as coined by VirtuoZo Systems) 
virtual photogrammetry. Scanned images are processed by 
specialised computer software to produce the required products, 
namely, digital maps. Such software, when bundled with the 
required computer, is collectively termed the Digital 
Photogrammetric Workstation or DPW. An obvious advantage in 
the DPW is the automatic measurement by computer instead of 
the eyes, hands and brain of the operator. As the technique of 
automatic object recognition within an image has not yet matured, 
the fully automated production of topographic maps is still limited 
to the production of orthophoto maps. 
During the evolution of photogrammetry to it's current digital 
state, a DPW, the Wuhan Digital Automatic Mapping System, or 
WuDAMS, initially proposed by Prof. Wang (Wang Z. 1978; 
Wang Z., 1995), was successfully developed at the Wuhan 
Technical University of Surveying and Mapping (WTUSM) of P. 
R. China (Zhang Z., 1986; Zhang J., 1993 and 1994). This DPW 
has now entered the market and is commercially known as 
VirtuoZo (the VirtuoZo Digital - or Virtual - Photogrammetry 
System). In this paper, the main functions and the theoretical 
foundation of VirtuoZo are introduced. These main functions 
include automatic interior and relative orientations, automatic 
measuring and transferring of the tie points for aerial 
triangulation, relative rectification to get the stereo model, image 
matching, interactive editing of the match results, automatic 
generation of digital terrain models (DTM), digital orthoimage 
generation, contouring, mosaic functions, landscape draping and 
semi automatic feature extraction. In the past fifteen years, 
countless theories and computational methods in digital 
photogrammetry have been developed, such as the theory and 
practice in one dimensional pixel arrangement along conjugate 
epipolar lines, the use of the bridging mode image matching 
method, global image matching by way of dynamic programming, 
probability relaxation and neural network techniques, etc. It is 
quite evident from the speed of VirtuoZo that one of our most 
significant achievements has been in solving the problem of image 
matching. 
424 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B2. Vienna 1996