The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
3.BRIEF ANALYSYS OF EXISTING CARTOGRAPHIC 
MODELS 
Before building an own cartographic data model, main existing 
cartographic model have been reviewed and examined; two of 
these are the Intesa-GIS Model (made in Italy) and the 
CityGML Model”, made in Germany. 
The first main model analyzed, called “Intesa GIS”, is a 
cartographic model built by an italian working group composed 
of several members of the main public italian bodies. 
Main features of the model are: 
a. geometrical schema is based on the ISO-19107 
standard, that includes solids; notwithstanding this, 
not every classes of the ISO-19107 standard are 
adopted, to easily distinguish between 2D and 3D 
features and to make easier the use of this model by 
unskilled users; 
b. generally it's impossible to directly define objects 
with three dimensions, or volumes; 
c. surface are represented in two dimensions only, but 
there is an “hybrid” way to model 3D surfaces, using 
a class of spatial attribute called “B3D surface”, that 
is constituted by a 2D surfaces joined with a 3D ring 
that is the boundary of the surface; 
d. topological constraints are defined separately from 
the geometry; 
e. compound, complex and aggregate objects are 
defined; 
f. levels of detail are not defined, because geometrical 
information is frequently and strictly linked to the 
scale of the representation; 
This cartographic model is not fully oriented to 3D 
visualization and analysis, that represent important features to 
accomplish typical GIS operations. 
The “CityGML” model is an information model suitable to 
represent cartographic and urban objects, mainly oriented to 
structure 3D models of cities, carried out by a consortium of 
several German bodies, both public and private. 
It encodes a multi-level representation of cities, also including 
elevation, vegetation, water bodies, city furniture and more. It 
uses levels of detail, that allow the use cartography at different 
scale and with different contents, and only one geometric- 
topological structure. Besides CityGML introduces textures into 
GML, representing the graphic appearance of objects and 
making realistic models suitable for virtual reality applications. 
At present CityGML have undertaken the course to become a 
real OGC standard. 
According to the goal mentioned above, after the analysis of 
these cartographic model, looking also to OS MasterMap GML, 
CityGML, NEN3610, basing the work upon international 
standards (like ISO and OGC rules) a three-dimensional 
numerical cartography model has been implemented. 
4.THE GIANT3D MODEL 
In order to allow advanced use of cartography in GIS and 
WEB-GIS, this model respects well-established rules of 
numerical cartography and the need of three-dimensional 
visualization and analysis of urban and out-of-town landscape. 
Due to the fact that a unitary data model - collecting semantic, 
geometric and topological features - is needed, the research has 
been addressed to take GML rules (version 3.X) in 3D 
geographic data structuring processes. GIANT3D Model is an 
application schema based on GML 3.1 profiles. 
In the proposed cartographic model entities are classified as 
‘topological’ objects; topological description of spatial 
relationships speeds up computational processes and allows the 
description of shared geometric features through simple 
procedures. 
During data collection and processing, therefore, it takes to 
verify the existence of topological exacteness in the geometric 
structure of entities, such as shared features and closed 
polygonals. 
In order to ensure the consistency of 3D data, some rules have 
been set: 
• objects’ footprints must be surveyed; 
• footprints are significant data in Digital Terrain 
Models processing; 
• duplication of elements connecting different objects 
(shared walls, lines belonging both to road and to 
building footprint, etc.) must be avoided; 
hidden objects (subways, porches, galleries, etc.) 
must be surveyed. 
Most of existing urban 3D models are suitable for visualization 
and navigation; objects are often modeled as simple geometric 
elements (i.e. boxes), whose surface is usually mapped with 
photos. Such models are not suitable for spatial analysis. 
The definition of main classes has been carried out according 
both to geometrical and thematic features; such classes have 
been used to define a structure of data that goes beyond the one 
commonly used in Italy -strictly organized in layers- based on 
the “recommendations of the Italian Geodetic Committee”. 
Feature classes are grouped -and linked to each other- into a 
schema that is split in different smaller schemas, according to 
the main classes described below: 
Real objects 
Modelled objects 
Building 
Building part, ground size, maximum size, 
crowning, terrain-building intersection, 
horizontal part, accessory perimeter, 
components of their bound (surfaces 3D) 
Artefact 
Facility, fence, single, terrain-artefact 
intersection, accessory perimeter 
Transportation 
infrastructure 
Vehicular area, additional area (road-bed, 
pedestrian areas, parking lots, etc.), terrain- 
transportation intersection, axis 
Infrastructure 
Platform, additional area, axis, terrain- 
infrastructure intersection, accessory 
perimeter 
Vegetation 
Accessory perimeter (open space), terrain- 
open space intersection, green area, 
additional green area, single plant 
Hydrographic 
entity 
Water sheet surface, terrain-water sheet 
intersection (outlines of rives and lakes) 
Terrain 
Under development 
Table 1. Classes of objects in GIANT3D model