ıbul 2004
team of
liar with
oduction
attribute
obtained
ge were
ypes of
capture
various
5111280
orland).
ailways,
ires and
cations,
Ss and
ly used
sector
ce the
nct and
ther or
image,
| scales
It was
ed for
uld be
catures
odland
\S may
lincar
fences
nush in
tic and
tween
s and
ture of
nd the
> that
ie for
could
nance
In a
more
ely to
iS this
| Dual carriageways
: Garage blocks
: Major sea defences
y
— 9^
; Tracks & paths
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B2. Istanbul 2004
should only be used as a secondary measure, to supplement the
information collected by other means.
Map scale
C14
EEE
= = =
S = =
7 : e = = eS
Features = = n
um C3 e
:2500
All vegetation
M
; Major landscape changes pans ;
[mand ae
Airports | y n n
Railways n n
; Non-residential buildings EPIS n n
Extensions to commercial buildings EE m ©
Water features Ea ^m 2 m 7
| Quarries ios m m
Housing & associated features n n
| Field EU : m - m
| Minor property boundaries m , m
n n
Major property boundaries
Telephone boxes In n n n n
Electricity transmission lines n: n n n n
Tide lines T nonem n n
Table 3: Analysis of the types of features which can be
identified from QuickBird imagery, at various
national mapping scales. Key: y = yes - feature can
be captured; n = no - feature cannot be successfully
captured; m = maybe - in some circumstances the
features can be captured, in others, not.
3. CHANGE DETECTION AND MONITORING
3.1 Change Intelligence
Although the main duty of a mapping agency is to update
geospatial data, such data cannot be updated unless it is known
where topographic change has taken place. Therefore, change
intelligence forms a very important part of the map revision
process. There are many different ways to identify change, one
of the most important being local observation by surveyors in
the field. Local planning authorities may provide planning
information, as do commercial change detection agencies. In
Great Britain, new housing development plans are often
supplied to mapping agencies by architects and house building
consortia.
To supplement this direct observation and notification of
change, there is a role for imagery. High resolution satellite
sensor imagery may allow surveyors to find areas of change
which would not be detected using the other methods. For
example, in areas undergoing continual change, such as central
London, satellite imagery could provide regular snapshots of
the area, enabling surveyors to constantly monitor and capture
topographic change.
In rural areas, the change intelligence requirements are often
different. Buildings may be constructed without planning
permission; field boundaries are changed from year to year;
hedges and woodlands may be removed, or newly planted.
These will often be in remote areas; and therefore do not come
to the attention of local surveyors or any of the change
notification bodies. In these areas, it is suggested that imagery
can prove a valuable tool for change e Ye ecially if
this use can be combined with a role as a source of data for the
subsequent capture of the topographic change.
To test this hypothesis, extracts of QuickBird images in the
Salisbury and Manchester areas were examined to detect
changes. These results were then compared with change
intelligence obtained by conventional means.
3.1.1 Change Intelligence Results
Previously unrecorded changes were detected in both the rural
and urban images. In the urban area of Manchester, most of the
changes were classed as “category À” (this category includes
new housing, commercial, industrial, community and public
sector buildings, roads, rail and other communications links).
In the rural area, most of the changes were “category B” (this
category includes small agricultural and horticultural buildings,
quarries and other surface workings, field boundaries, water
features, vegetation features, tracks and paths). In Manchester
the analysis uncovered an average density of one site of signifi-
cant change per square km; the corresponding figure for Salis-
bury was 0.66 sites per square km. Note that these were
changes which had not been identified using traditional change
intelligence techniques.
The main types of change identified in the Manchester study
site were: building demolitions (industrial and housing); newly
built industrial units; railway demolitions and minor road al-
terations. It was not possible to detect changes such as new
traffic calming measures; small property boundary changes; or
mobile-home movements. Of course there are other changes
which are impossible to detect using any type of imagery, in-
cluding name changes, conversions of buildings from agricul-
tural to residential, or address changes.
In the Salisbury area, the main types of change were to typi-
cally rural features such as fences, tracks and vegetation
boundaries. Although recorded as *Category B" and therefore
regarded as slightly less important to the large scales data col-
lector, these features are of significant interest to the leisure
map user and are therefore important to the small scales map
update process. As in the large-scale case, there are many
changes which cannot be observed from imagery alone, includ-
ing non-topographic data such as tourist information.
These results indicate that QuickBird imagery can be used to
identify topographic changes for both large- and small-scale
mapping. The cost of the imagery may well be the sticking
point. At current costs, it would not be economically viable to
use QuickBird data (or any other high resolution satellite sensor
imagery) solely for the purpose of change intelligence. 1f, how-
749
