4
Progressive Techniques for Efficient Vector Map Data
Transmission: An Overview
Michela Bertolotto
University College Dublin, Dublin (Ireland)
4.1 Introduction
Progressive data transmission techniques are commonly used for data exchange over
the Internet: a subset of the data is sent first and then incrementally refined by subsequent stages. The advantages of progressive transmission have been highlighted by
many researchers [5–8, 10, 22] and include efficient data transmission (as smaller
files are sent), quick response, and, possibly, transmission of only the most relevant
detail. Besides being able to perform preliminary operations on temporary versions
of the data, during progressive transmission, users can realize that the detail of the
currently displayed representation is good enough for their purpose and so can decide
to interrupt the downloading of more detailed representations (stored in larger files).
Therefore, both time and disk space can be saved. This is particularly important for
users in the field trying to download datasets of interest. Even if the fully detailed
version is needed, they can start working with a coarser version while more detail is
being added progressively.
The first and most successful implementations have been developed for raster
images [12, 21, 41]. The data is efficiently compressed with different techniques and
sent to the user. The full resolution image is reconstructed on the user’s machine by
gradually adding detail to coarser versions. The success of progressive raster transmission relies on the availability of effective compression techniques for such data:
these techniques provide good compression ratios while causing low information
loss. Furthermore, they are efficient and relatively easy to implement.
These mechanisms are applied also in the spatial domain: raster geospatial
datasets (including high resolution satellite images, aerial photos, and scanned maps)
can be exchanged progressively from a server to a client. This is acceptable for
applications that mainly involve visualization. However, a raster version of the
data might not be adequate in certain applications, including contexts in which
actual object manipulation is involved. In these cases a vector representation of
the data is required. Vector data sets consist of collections of spatial entities in the
form of points, polylines, and polygons that are related through spatial relations
(e.g. topological, metric, and direction relations [20]). Examples include thematic
maps, road network maps, city maps, mesh-based digital terrain models, and so on.
Progressive Techniques for Efficient Vector Map Data
Transmission: An Overview
Michela Bertolotto
University College Dublin, Dublin (Ireland)
4.1 Introduction
Progressive data transmission techniques are commonly used for data exchange over
the Internet: a subset of the data is sent first and then incrementally refined by subsequent stages. The advantages of progressive transmission have been highlighted by
many researchers [5–8, 10, 22] and include efficient data transmission (as smaller
files are sent), quick response, and, possibly, transmission of only the most relevant
detail. Besides being able to perform preliminary operations on temporary versions
of the data, during progressive transmission, users can realize that the detail of the
currently displayed representation is good enough for their purpose and so can decide
to interrupt the downloading of more detailed representations (stored in larger files).
Therefore, both time and disk space can be saved. This is particularly important for
users in the field trying to download datasets of interest. Even if the fully detailed
version is needed, they can start working with a coarser version while more detail is
being added progressively.
The first and most successful implementations have been developed for raster
images [12, 21, 41]. The data is efficiently compressed with different techniques and
sent to the user. The full resolution image is reconstructed on the user’s machine by
gradually adding detail to coarser versions. The success of progressive raster transmission relies on the availability of effective compression techniques for such data:
these techniques provide good compression ratios while causing low information
loss. Furthermore, they are efficient and relatively easy to implement.
These mechanisms are applied also in the spatial domain: raster geospatial
datasets (including high resolution satellite images, aerial photos, and scanned maps)
can be exchanged progressively from a server to a client. This is acceptable for
applications that mainly involve visualization. However, a raster version of the
data might not be adequate in certain applications, including contexts in which
actual object manipulation is involved. In these cases a vector representation of
the data is required. Vector data sets consist of collections of spatial entities in the
form of points, polylines, and polygons that are related through spatial relations
(e.g. topological, metric, and direction relations [20]). Examples include thematic
maps, road network maps, city maps, mesh-based digital terrain models, and so on.