An application system and its
intelligence for earth resource satellite
Chen Shupeng Li
Zhirong Chinese Academy of Sciences
Wang Yuan
Chinese Academy of Space Technology
Abstract
An application system is proposed for the Earth Resource Satellite
(ERS) project. The system under consideration consists of five subsystems:
- The data receiving and recording subsystems;
- The data preprocessing and production subsystems;
- The applied analysis and specialist subsystem (AAS);
- The data simulation and evaluation subsystem;
- The data distribution and user's service subsystem.
Except
the traditionally satellite receiving and preprocessing subsystem, the
important subsystem of AAS will be discussed in detail. It is basically
geographic information of remote sensing imagery.
Such a subsystem
will provide the real time and effective dynamic data information to
various users for different economic constructions. It is also able to
supply the global static information and thematic map of a certain part of
the country or even the whole country.
Earth resource satellite
(ERS) needs an application system Since the first American ERST
satellite was successfully launched in 1972, its data was used in more
than 150 countries and its social and economical benefits were quire
obvious. Btu the use of satellite data for remote sensing applications in
different countries were separated and less concerned for whole country
wide, so the results of applications only solved some regional problems.
if we want to make the most of ERS works, a complete application system
should be considered. This application system is also one of the five
parts of the whole satellite engineering. The other four parts are
satellite, rocket, control and communication and launching field. The
relations between then are shown in Figure 1. As an application system, it
has the capability to research renewal resources in an large area and
dynamic monitor the environment.
Fig. 1 Earth Resource Satellite and
Applications System In the future, the application system of ERS
can provide global thematic maps or statistics with practical values for
national planning or decision making.
For operational process, the
auxiliary data will be input to the system at different stages and the
system also output different products at different stages to satisfy the
demanding of different uses. Fig.2 shows the flowchart.
Fig. 2 Information flowchart of
application system and It's feedback The conception design
and technical bases of subsystems
- Receiving and preprocessing subsystems
This systems is the
data source of the whole application system. It undertakes the data
receiving, recording, radiometric and system geometric corrections, and
the standard films and computer compatible tapes (CCT) productions. The
subsystem should be an operational system and has the capability of mass
production for the requirement of dynamic monitor.
To cover the
land and sea of whole country in a period of 1-2 years and renew the
environment data and maps of renewable resource require high through out
of ground station. For example, the total landarea of China is 9.6*106
Km2. It needs 85 international map sheets for whole territory and 65
sheets for land areas. So far the standard products of ground station
are generating geocoded imagery which is which is mission-independent
geocoded remote sensing products and involves projection onto a fixed
and suitable common cartographic projection.
The geocoded image
products are based on map sheet and easy to use with existing map
referenced data.
Neither the high throughput nor the format
change of products are more higher requirements to the data processing
of future ground station.
- Applied analysis subsystem:
This is an important subsystems
in the application system. The products from the preprocessing subsystem
will be processed and transformed to useful thematic maps and statistics
information in this subsystem. The effectiveness of satellite data will
be checked in this subsystem also. In order to output different maps and
make decision quickly, the subsystem not only has powerful processing
capability but also combine the specialist knowledge an human
intelligence to realize high precise quantitative analysis and
systematic intelligence.
The total subsystem consists of four
parts:
- Digital image processing,
- Geographical information system
- Applied analysis model, (discussed in next articles)
- Computer aided mapping
Fig 3 illustrates the diagram
of the subsystem.
- Image processing
The image processing will comprise two
aspects. One is general image processing which includes contract
enhancement, edge detection, statistical calculation, information
extraction and image classification. The other aspect is using the
digital terrain model (DTM) or ground control points to do the high
precise geometrical correction, large scale mosaic, graphic data
handling, and the map output through the plotter.
Special
attention should be paid to the equipments of digitized graphic input
and thematic map output.
Fig3. Information Flowchart of
Application System
- Geographical information system (GIS) and computer thematic
mapping
The new trend in the remote sensing field is the
development of the image processing system with GIS. In the past
years, the image processing system and GIs system have developed
independently and parallels. An important problem to be solved for
final convergence of two systems is the interface between these two
systems.
The main purpose of the GIS is to compensate the
shortage of information in the Earth resource satellite application
system. The data contents of GIS can be divided into two types:
- Graphic data base, databases
It consists of map database,
Digital terrain model and Database of place.
- Statistics data base : this data base had the items of space
coordinate on time sequence. There are:
Geophysical field,
earth surface observation, natural environment base and social
economic base. From the software point of view, the main question is
the management system of spatial database and the correspondence
relational and geometrical data base. The special
software for GIS, compared with generation image processing, has a lot
of graphic processing functions. It will consists of following items:
- Digitization of graphic and map: The determination of the value
and direction of the graphic, the digitization of the point, lien
and polygon etc…
- Graphical edition: Processing of the cross point, joint lines
between the gap of two lines, registration and linking of different
graphical files, converting the data format between raster data and
vector data.
- Graphic processing: Curve spline, putting the value on the
isopleth and 3-dimension display.
- Graphic output: Class: field map, isopleth map, statistics map,
symbol and annotation….
Due to a large amount of
information in each map sheet, it requires large output size of he
map, say 14000 * 800mm
Besides, different thematic maps have
different map scale. All of these need a high throughput of the CAC
system.
The relations between image processing, GIS, and CC
are shown fig. 4.
- The data simulation and evaluation subsystem:
This subsystem
undertakes the data simulation experiment before and later the launching
satellite, such as:
- Pre-launching airborne data simulation research, it will include
the selection and modification of spectral band of earth resource
satellite.
- Check the availability of the simulated sensors in the
applications of the crops forecasting, landuse, geologic exploration
and water survey.
- Post-launching evaluation f the technical specification of the
satellite. It will include the accuracy of geometry, ground
resolution, and radiometric calibration of sensors.
- Data distribution ad user's service subsystems:
The products
of application system it will be supplied to whole world. So it not only
need a daily distributive system, but also asks to organize special
symposium, publishing bulletin and introducing the experiments of
products application to the users.
The subsystem could register
and log the processed image to build a directory of acquired imagery and
link it to archive inventory. The quick look ask copy of received image
are provided in this subsystem also. The intelligence of
application system It is well known that the interpretation of
remote sensing imagery by specialist knowledge is much better than the
automatic classification by using most advanced computers. The image
processing system in combination with GIS will increase the accuracy of
the classifications other than only spectral information in common image
processing system. If this system incorporation with specialist knowledges
(i.e. human intelligence), the advantages of this new system are quire
evident. The major superiority of the new system over the common image
processing and GIS is its incorporation with human brain in the process of
image interpretation or classification. Due to using the human
intelligence in pattern recognition and applied analytic model, many
practical problems in the economic construction can be solved more quickly
and accurately. Fig. 5
Fig. 5 Method and Symbol of Image
processing
- Contents of Intelligence in the application system
To
establish different analysis model and softwares by the specialist
knowledge are important task in the application system's intelligence.
This will promote the system to a expert system and intelligent machine.
As for the analysis mode's it should include following six
areas:
- Renew the national basic topographic map and data base of GIS;
(1:250,000 - 1,200,000) scale)
- Monitor land resource; (desertification, urbanization etc)
- Qualitative evaluation and estimation of water resources;
- Estimation products of main crops for large area; (wheat, rice,
cotton…)
- Monitor of the flood, drought and other disaster in a large
area;
- Geological mapping and exploration of the mineral resources in
different regions.
Beside the analytic model, the
intelligences of image and graph processing are necessary, such as map,
digitizing interpretation.
- The technical methods for establishing applied analytic model.
- First review and evaluate the works of former analytic and
software system:
- Establish the digital formula and the procedures of logical
inference
- Digital analog that combine with the physical model and practical
conditions
- Design the software package that reflect the difference between
each district area and introduce the real corrective
coefficients
- Design each special intelligent machine.
The concept
of the classifiable machine should add to the above procedures. Several
special analytic knowledge databases are needed to enhance the man
machine interaction. The fuzzy mathematics and pattern reorganization
are used to perfect the logical interference model. See Fig 6.
Conclusion
- The Earth resource satellite has developed more than 20 years. S for
sensors and platform of satellite also got much progress. But the
information processing and the applications of satellite data lagged
behind others in the whole Earth Resource satellite system. In order to
raise the application's level, it should have a general plan to
establish an operational system that combines the information
processing, analytic model, GIS, and user's service as a whole.
- Due to the mass volume of the data in application system, the
processing speed must be quick and fast. In the application side, it
needs to solve the practical problems and qualitative analysis. All of
these require high qualitative products and system's intelligence. Then
the operational application system had the capability to solve the
practical problems such as the real time monitor of landuse or disaster
for whole national wide.
- The intelligence of application system includes many sciences, such
as space science, information science, geo-sciences and bio-sciences. It
also interacts each other among the computer, remote sensing,
information processing, and automation. So the development of the
intelligence of the system will open the new areas of the high
technology.
References
- Wang Da-Heng & Pan Hou-Ren, Acad. Sinica, Presented at ISY
SAFISY Conference, UK Feb. 1989.
- Chen Shu-Peng, Presented tot eh ISY SAFISY conference, UK. Feb.
1989
- Chen Shu-Peng, Selection from the bulletin of the Chinese Academy of
Sciences, 3, 1987 47-53.
|