169 
A NEW CONCEPT: THE SOLID IMAGE 
Leandro Bomaz ( # ), Sergio Dequal ( # ) 
C) Politecnico di Torino - Dipartimento di Georisorse e Territorio 
C.so Duca degli Abruzzi, 24- 10129 Torino 
Tel. +39.011.564.7687 / 7601 Fax. +39.011.564.7699 
Email: leandro.bornaz@polito.it: sercio.dequal@nolito.it 
KEY WORDS: Digital photogrammetry, Analysis, Image, Application, Architectural Heritage, 3D image, laser scanning. 
ABSTRACT: 
A photo image can be considered a central perspective of the acquired object with good approximation. If the internal and external 
orientation of the camera are known, in case of a digital image it is possible to establish the direction in the space of each object point 
represented by a pixel in the image. If only one image is available, it is impossible to determine the spatial X,Y,Z position of such 
object points, because the simple direction is insufficient: all the points along that direction would give the same image point. 
By means of a dense digital elevation model (DDEM) of the acquired object, every pixel (and therefore every direction in the space) 
can be associated to the value of distance between the centre of perspective and the object point represented by the pixel itself. In this 
way each pixel can be referred to the 3D position of the corresponding object point in an absolute reference system. 
The DDEM can be derived from an existing map or by the use of surveying instruments and procedures. Recently, a series of new 
instruments, based on the laser technology, have been introduced on the market, giving the possibility to obtain a DDEM in a quick 
and cheap way. 
The integration of the DDEM geometric data with the image radiometric data suggests a new concept: the “solid image”. 
Definition, first results and applications of the solid image was presented in the ISPRS Comm..V, WG V/4 meeting in Ancona, July 
2003. In addition, this paper describes the very recent further developments and the future activities on this field. 
. INTRODUCTION 
An image can be considered a central perspective of the 
photographed object. If internal and external orientation are 
known it is possible, for every pixel of the image, to 
determine the direction of its projecting ray in the space. If 
only one image is available, that direction is insufficient to 
reconstruct the 3D position of the object point (i.e. its X, Y 
and Z coordinates). In fact, all the points positioned along 
that direction in the space would give the same pixel on the 
image. 
The recent introduction in the market of laser scanner sensors 
in the field of survey instruments allows dense digital 
elevation models (DDEM) of the object to be obtained. 
The integration of the laser scanner data with the digital 
image data can give the possibility of associating a value of 
distance to every direction in the space, defined by each 
pixel. This distance is calculated as the distance from the 
perspective central point (the “taking point”) and the object 
point represented by the pixel. If two angles defining a 
direction in the space and the above mentioned distance are 
known (these 3 figures are called “spherical” or “3D polar” 
co-ordinates), it is easy to reconstruct the spatial position (i.e. 
the X,Y,Z co-ordinates) of every point represented in the 
image. 
The traditional RGB radiometric data of the image together 
with a 3D model acquired by the laser scanner lead to a new 
concept: the “solid image”. 
Advantage and potentials of this innovative product are due 
to the possibility of: 
getting the spatial position of points in the object 
reference system; 
direct and easy carrying out of correct 3D 
measurements on the image; 
obtaining a great deal of information in a simple and 
rapid way, using the high qiality original images and 
any traditional photo viewer software available on the 
market. 
Based on the solid image, it is possible to foresee several new 
applications: a second image of a stereoscopic pair can be 
created, a RGB coloured virtual 3D model can be carried out, 
etc. 
2. THE SOLID IMAGE 
2.1 Definition 
A “true colour” digital image is made up of 3 matrixes, 
named R, G and B, each one containing one of the 3 
fundamental colours (Red, Green or Blue). 
object space 
Figure 1. Relationship between image space and object space 
The solid image is defined as a classical digital image that, in 
addition, records the distance values of the represented points 
from the taking centre. Assuming that the digital image is a 
central perspective of the photographed object and the 
internal orientation parameters of the image are known, if the