Base map revision using high
resolution satellite data
Col. Somkiat
Aisanont Royal Thai Survey Department, Bangkok 10200,
Thailand
Abstract Maps are essential
database for both planning and fieldwork operation. Up-to-date maps are
necessary because obsolete maps cause users to be confused and create
problems during work.
The conventional methods of map revision
with aerial photographs are so costly and highly time-consuming that base
map revision cycle cannot be done as planned. However, the recent
availability of high resolution satellite data such as SPOT offer an
interesting alternative for map producing and updating. The Royal Thai
Survey Department, which is responsible for the production of topographic
base maps for both military and general purposes realizes the usefulness
of satellite data, hence initiating this project.
The objective of
the project was to update base maps using SPOT imagery, Project area
covered 24 map sheets of 1:50,000 located in the Central plain where it is
dominantly flat terrain. SPOT panchromatic level 2 images were used, as
main data source to add new information or changes of features on the base
maps and multispectral level 1B were provided as supplementary
information. Topographic base maps at 1:50,000 and 1:250,000 scales were
simultaneously revised using optical transfer device.
Introduction The Topographic maps which cover the whole
area of Thailand are at the scales of 1:50,000 and 1:250,000. The latter
series was derived from the first one.
The series of 1:50,000
which has been used widely, comprises 130 map sheets for the whole
country. This series was mostly produced in 1969 and only some map sheets
have been revised.
The Central plain Area is predominantly a flat
terrain and has been developed rapidly which causes the map to be out of
date. The Royal Thai Survey Department-RTSD is responsible for topographic
mp production, revision and provision for both military and civil
supposes. With a view to find an appropriate technique for map revision,
the department has therefore put an effort in conducting the zloty project
using satellite imagery for map revision. The high-resolution satellite
images used for such purpose are available from the Thailand Ground
Receiving Station, which has high capability in creiving Landsat TM, SPOT
and MOS-1 data.
Objectives
- This project aims at using high resolution satellite imagery for
updating and adding new features on to base maps at the scale of
1:50,000. A total of 24 map sheets will be updated and printed for non -
permanent use.
- The products of item 1 will then be used for map revision at the
scale of 1:250,000.
- The updated maps can be utilized as an input for other cartographic
purposes.
Target Area The target area of the project
was selected by the RTSD Committee. The location of the area is in the
Central Plain of Thailand where map details have been so much changed. It
corresponds to sheets no. ND 47-7 of 1:250,000 scale and covers a total of
24 map sheets of 1:50,000 scale which are 4837 I, II, III Iv 4838 I, II,
III, IV 4937 I, II, III, IV 4938 I, II, III, IV 5037 I, II, III, IV and
5038 I , II, III, IV.
Methodology Several methods have
been available for map revision. One should also take into account the
development trends of this technology so as to choose a proper means
suitable for the existing mapping system that one currently has.
The methodology used in this project is an economical one and can
be performed directly onto paper maps, which is the current forms of maps
for Thailand, The updating is carried out by using an optical transfer
device called Procom-2 as the following steps.
- Selection of maps and images
- it seems beneficial to update maps that have never been revised
since the first edition. The type of terrain is also another important
element. Flat terrain should be selected to preclude relief
displacement. The target area of this project covers 24 map sheets of
1:50,000 (Table1) located in the flat terrain of the Central Plain.
Information on the existing base maps was collected in 1969. Since one
map sheet of 1:250,000 scales comprises 24 map sheets of 1:50,000 the
updating of these 24 map sheets would then lead to updating of the
1:250,000 map sheet number RND 47-7.
- SPOT images, both Panchromatic Linear Array (PLA) and
Multispectral Linear Array (MLA), in the form of positive films were
used.
PLA images, which provide a resolution of 10 m X 10 m,
were essentially used for interpretation and delineation of new
features. Therefore, images of level 2 (Geocoded, precision corrected)
were selected. Such images were also of vertical or near vertical
viewing angle, cloud free and recent acquisition. The index of a total
of 34 PLA level images used for updating is shown in Figure 1, which
the details of original PLA images are shown in Table2.
MLA
images which provide a resolution of 20 m x 20m were used as
supplementary data for interpretation of some features like present
forest boundary, encroachment area that has become agricultural
Landuse of urban area, new water bodies such as irrigation canals,
ponds, reservoirs, Therefore MLA images of level 1B (full scene, bulk
corrected) were used and the details are shown in Table 3.
- Interpretation of SPOT images
Interpretation of SPOT
images were conducted by using an article transfer device called
PROCOM-2 . With the device, the image alas could be enlarged and
projected on to 1:50,000 maps. Because of is zooming and roaming
capabilities, the PROCOM-2 allows super imposition of the satellite
image onto the map and working directly at the map scale.
Common
features on the map and image such as roads, rivers, creams, ponds or
reservoirs, etc. were used for local registration in such sub-area of
the base map. All types of features were interpreted and drawn on the
base map with pencil.
- Field verification
All features that were located onto the
maps in the office are then checked in the field. The purposes of field
work were to verify and classify the detail of new features such as type
of roads (hard or loose surfaces), road number and width, type of
agricultural and, names of features such as school, monastery, village,
reservoir etc.
- Post field verification
The ambiguous features from
interpretation in the office could be identified after field
verification. Some features could not a detected by satellite images
like important official buildings then were covered with tree crowns.
Such features could be mapped fleld verification. Field verification is
by all means necessary and very useful.
- Revision and colour separation of the originals
After
field checks, the new features and changes were drawn overlays. These
overlays were used for scribing on the scribing rats derived from the
originals of existing base maps.
Colour separation was generated
on positive films. All of the revised positive films were finally
integrated to generate new negative originals for printing in the next
step. However, before the process of printing, colour proof had to be
performed to check for me errors.
- Colour separation for 1:250,000 map compilation
After 24
map sheets at the scale of 1:50,000 were updated, of the revised
originals were transferred to 1:250,000 scale.
- Printing
- A total of 24 updated maps at 1:50,000 scale wee printed the
standard legend. The remark non-permanent use is shown on top the
sheet. Lower left corner denotes Updated for all major sturesusing
SPOT satellite imagery otained in.........
- One sheet of updated map at 1:250, 000 scale was printed is sheet
number is ND 47-7.
Results The project area is mainly characterized by
flat terrains located in the Central Plain. Base maps of this project have
never been revised since the first edition in 1969. Both new features and
the changes could be added onto the existing base maps such as roads.
Irrigation canals, reservoirs, deforestation area that has become
agricultural land and residential area. The utilization of high-resolution
satellite images has proved to be a very effective means for map updating
for this area and could provide map users with more up-to-date and useful
information.
At the time of writing (August 1990), RTSD has
assigned ground survey party to evaluate positional mapping accuracies.
Conclusions and Recommendations In general, aerial
photographs are primary data for as revision. Satellite data will be
utilized in particular cases such as when information on the maps is out
of date or when there are is urgent requirements from the users. It is
most beneficial to use both aerial photographs and satellite images in a
complementary approach.
Methodology for updating has to be based
on the existing equipment and other resources. However, it should be
pointed out that remote sensing technology is being developed rapidly with
high competency. Investment for expensive equipment should be considered
carefully in term of benefits. Simple methods using expensive equipment
such as visual interpretation by an experienced interprets and appropriate
type of images can also provide good results.
Acknowledgements The author wishes to acknowledge RTSD
colleagues for that consistent cooperation. My thanks are also extended to
the Remotes Sensing Division of National Research Council of Thailand for
the support provided in this project.
References
- Turner, A.M. and D.R. Stafford, (1986) Operational Revision of
National Topographic Maps Using Landsat Images. The Canadian Experience.
Proceedings of the Commission it Symposium on Mapping & Modern
Imagery, ISPRS vol. 26, part 4, Edinburgh.
- Gregory, A.F. and Moore, H.D. (1986) Economical Maintenance of
National Topographic Data Base Using Landsat Images Photogrammetric
engineering and remote sensing vol. 52 No. p 519-524.
- Fleming, E.A. (1982) Topographic Map Revision Using Satellites
Imagery. Prepared for the second National Workshop is Engineering
Applications of Remote Sensing, Edmontes Alberta.
- Aisanont, Col. S (1988) Topographic Map Updating Using SPOT Imagery
Abstracts of the Mid-Project Seminar under the CIDA-Thailand Landsat
Project, January, Thailand.
- Aisanont, Col. S (1988) Base Map Updating in Thailand Using SPOT
Satellite Imagery. Proceedings of the Ninth Asl Conference on Remote
Sensing Vol. 7 Session D.
Figure 1 Panchromatic, level 2
(Geocoded, Precision corrected) SPOT images used.
Table 1 Information of 24 base map.
No |
1850,000 BASE MAPS |
EXISTING MAP DATA (YEAR) |
SHEET NUMBER |
SHEET NAME |
1
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 |
4837 i 4837 ii 4837 iii 4837 iv 4838 i 4838
ii 4838 iii 4838 iv 4837 i 4837 ii 4837 iii 4837
iv 4838 i 4838 ii 4838 iii 4838 iv 5037 i 5037
ii 5037 iii 5037 iv 5038 i 5038 ii 5038 iii 5038
iv
|
BAN LAM HOEI BAN LAT YA AMPHOE SAI YOK BAN KANG
RIANG KHAO PHU TOEI BAN NONG PRU AMPHOE SI SAWAT BAN
PHUNG AMPHOE U THONG BAN THUNG KHOK CHANGWAT
KANCHANABURI AMPHOE BO PHLOI BAN NONG KRATHUM BAN NONG
CHOK BAN KRAP BAN NA TA PIN AMPHOE PRAK HAI AMPHOE
LAT LUM KAEO AMPHOE BANG LEN HANNGWAT SUPHAN
BURI CHANGWAT SING BURI CHANGWAT ANG THONG AMPHOE SI
PRACHAN AMPHOE DOEMBANG NANGBUAT |
1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1969 1974 1969 1969 1969
| Table
2 original Panchromatic SPOT images.
NO. |
K - J |
SENSOR |
ACQUIRED |
VIEW ANGLE |
1 2 3 4 5 6 7 8 9 10 |
259 - 320 259 - 321 260 - 320 260 - 321 260 -
322 261 - 320 261 - 321 261 - 322 262 - 321 262
- 322 |
HRV 2 HRV 2 HRV 2 HRV 2 HRV 2 HRV1
HRV1 HRV 2 HRV1 HRV 2 |
1989-05-08 1988-07-10
1988-12-24 1988-12-24 1988-12-24 1988-12-24
1988-12-24 1988-12-24 1989-01-19 1988-12-19 |
-1.94 13.10 -5.95 -5.95 5.95 -2.30 -230
-595 1.94 -5.95 | Table
3 Multispectral SPOT images, level 1B (full scene, bulk
corrected)used.
NO. |
K - J |
SENSOR |
ACQUIRED |
VIEW ANGLE |
1 2 3 4 5 6 7 8 9 10 11 |
259 - 321 259 - 322 260 - 320 260 - 321 260 -
322 260 - 320 261 - 321 261 - 322 262 - 320
262 - 321 262 - 322 |
HRV 2 HRV 2 HRV 1 HRV 1 HRV 1 HRV 2 HRV 2
HRV 2 HRV 2 HRV 2 HRV 2 |
1988-11-12 1988-11-12 1989-05-08 1989-05-08 1989-05-08 1989-01-19 1989-01-19 1989-01-19 1988-12-03 1988-12-03 1988-12-13 |
6.32 6.32 2.60 2.60 2.60 -1.31 -1.31 -1.31 10.38 10.38 24.77 |
|