154 
Commission II ISPRS Congress 1984 
Instrumentation for topographic mapping from satellite data 
Ian Dowman 
University College London 
Introduction 
The past ten years have seen a great increase in the use of data derived 
from sensors on board satellites for topographic mapping; many papers have 
been written which investigate the geometric conditions and the accuracy 
which is to be expected. So far very little data with real potential has 
been produced and little attention has been paid to the methods of plotting, 
particularly when using stereoscopic images. Only certain types of image 
can be used in conventional plotting instruments and the problems of 
using digital data have barely been tackled at all. Some proposals have 
been made for automatic mapping systems but there are questions hanging 
over the implementation of such systems which must be answered. 
This paper is concerned with topographic mapping of plan and height at 
1:100 000 scale or larger. The paper will not therefore consider Landsat 
data but will confine itself to photographic systems, push broom systems 
and briefly, with radar. After an initial section on the data, discussion 
will centre on instrumentation which is relevant to the data. 
Sensors 
The characteristics of the most important sensors are given in Table 1. 
The primary form of data from a photographic system is a negative which 
contains all the information recorded of a single scene, by the camera. 
The main variables when considering methods of plotting are the principal 
distance of the camera and the format size. If the oblique photographs 
are involved then the amount of tilt is relevant. 
The geometry of a camera is characterised by the collinearity equations 
which represent straight lines joining points on the surface of the 
earth to the image points on the negative. A perspective bundle of such 
lines may be reconstructed in an analogue instrument by a mechanical rod 
or height rays and analytically by the collinearity equations. Any 
deviations of the imaging ray from a straight line must be accounted for. 
As a complete scene is imaged at one time, any reconstruction is 
independent of time. 
At present there are three photographic systems which should be considered 
as indicated in table 1. The metric camera on Spacelab is described by 
Schroeder (1982), Ducher (1980), the large format camera by Doyle (1982) 
and the MKF-6m by Zickler (1977). 
In March 1984 photography is available from the metric camera, the large 
format camera will be used during 1984. It is expected that both these 
cameras will be used again in future years. The MKF camera has been used 
on several Soyuz missions and on the Salyut space laboratory. 
The push broom systems are expected to become the standard satellites for 
continuous earth observation giving high resolution, stereoscopic cover. 
The collinearity equations also represent the geometry of a push broom 
sensor but the position and attitude parameters are constant only for 
recording a single line of data, thus to determine them with the aid of