GISdevelopment.net ---> AARS ---> ACRS 1990 ---> Poster Session

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The Central plain Area is predominantly a flat terrain and has been developed rapidly which causes the map to be out of date.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
  1. 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.

  2. The products of item 1 will then be used for map revision at the scale of 1:250,000.

  3. 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.
  1. Selection of maps and images

    1. 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.

    2. 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.

  2. 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.

  3. 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.

  4. 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.

  5. 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.

  6. 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.

  7. Printing

    1. 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.........

    2. 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
    1. 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.

    2. 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.

    3. Fleming, E.A. (1982) Topographic Map Revision Using Satellites Imagery. Prepared for the second National Workshop is Engineering Applications of Remote Sensing, Edmontes Alberta.

    4. Aisanont, Col. S (1988) Topographic Map Updating Using SPOT Imagery Abstracts of the Mid-Project Seminar under the CIDA-Thailand Landsat Project, January, Thailand.

    5. 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