Home / Infrastructures / Infrastructures / Image Processing System

InfrastructuresImage Processing System

Cooperating with domestic companies to develop an Image Processing System (IPS) is one of the important R & D programs led by the Taiwan Space Agency.

The main function of the satellite image processing center is for FORMOSAT satellite family image production, including image acquisition and scheduling, image data acquisition, image radiometric and geometric correction, image management, quality monitoring; and promoting the FORMOSAT satellite images through developing application platform as well.


Image Processing System and Demostration Center

The System Architecture of Image Processing System

Besides supporting smooth operations for the FORMOSAT-2/FORMOSAT-5 programs, the IPS team is currently developing an Image Processing System for the FORMOSAT-8 program. Save advanced features such as better spatial resolution and revisit frequency, the FORMOSAT-8 program is distinguished from the FORMOSAT-2/FORMOSAT-5 in two aspects: (1) in addition to linear sensor, the FORMOSAT-8 RSI is equipped with a frame sensor, (2) the FORMOSAT-8 program is planned to deploy a constellation composed of six satellites. The development of FORMOSAT-8 IPS follows strictly software system engineering process. According to Program Review Handbook issued by the TASA, each milestone is peer reviewed to assess the development status against targets and requirements. Figure 2 shows the software architecture of the FORMOSAT-8 IPS. The FORMOSAT-8 IPS consists of five subsystems: Planning and Scheduling Subsystem (PSS), Data Ingestion Subsystem (DIS), Data Management Subsystem (DMS), Data Processing Subsystem (DPS), and Image Quality Subsystem (IQS). Following sections describe briefly the functions of each subsystem.

The System Architecture of Image Processing System

One of characteristics of PSS design is asynchronous imaging mode. It provides the capability to enables sensor to scan over the earth’s surface and its direction is not limited to the direction of the ground track. In case of taking image along the coastline, it is expected to reduce the time to accomplish the task by performing only one acquisition instead of multi-acquisition.

Examples of imaging along the coastline

Establish the capability of absolute radiometric calibration of FS5 RSI, and assess inflight performance of the CMOS sensor



Data Ingestion Subsystem, DIS

Self-developed Front End Processing System (Data Ingestion Subsystem, DIS) to remove Black Box restrictions

The FORMOSAT-2 DIS is purchased from France Satellite Vendor because of the black box of the compression and encryption algorithms.

Data Management Subsystem, DMS
Data Management Subsystem (DMS) provides the IPS an operation friendly framework. Its functions are multifold: (1) maintains an archive of RSI data acquired from different RSI missions, (2) manages various data flows between IPS subsystems, (3) oversees the status and the activities of IPS subsystems, and (4) provides user interfaces for placing various product requests and monitoring IPS overall activities and system health. Utilizing up-to-date Information Technologies (IT) is essential for the development of the DMS. Taking advantages of well-established IT industry in Taiwan, domestic IT companies are invited to team up with members from the TASA for the development works. The TASA carries out the baseline design and details accordingly the system requirements and specifications. The team efforts from participant parties are planned for completing system implementation and validation works.

The FORMOSAT-5 DIS is self-developed due to self-made Satellite Hardware, and has the core technology (Decompression, Decryption and Source Format Structure) by our own.

After FORMOSAT-5 launch successfully, the FORMOSAT-5 DIS had been proved and verified the functionality and stability to do the daily operation.

Data Processing Subsystem, DPS

Data Processing Subsystem (DPS) generates standard image products for end users. The DPS performs key pre-processing computations which include radiometric equalization and geometric rectification operations. In addition, the DPS attaches to each standard image product an auxiliary data set containing information usually required for various remote sensing applications. Based on the needs of his or her applications, an end user may request a level 1A, level 2, or level 4 pre-processed image product. The definition of these pre-processing level is well defined and documented, each of them are tailored to fit the needs of a number of application fields. The DPS is in-house developed, which provides IPS team the flexibility of improving image quality and introducing new products.

Image Quality Subsystem, IQS

Image Quality Subsystem (IQS) is responsible for evaluating the RSI in orbit performance and maintaining for each RSI mission a global imagery gain setting file and a series of Calibration Parameter Files (CPF). In addition, the IQS examines the quality of image product before its delivering. During the mission commissioning phase, the IQS estimates Ground Sample Distance (GSD), Signal-to-Noise Ratio (SNR), and Modulation Transfer Function (MTF) to verify the RSI performance against the mission specifications. The annual SNR and MTF measurements are used for RSI health trend analysis. To optimize image dynamic range, the global gain setting file is referenced daily for each tasking activity. The CPF contains parameters characterizing geometric and radiometric behaviors of the RSI, which is essential for image product generation process. For sustainable supporting of current and future RSI missions, the TASA has organized a tasking force for the development of the IQS.

Data Cube

Promoted by CEOS (Committee on Earth Observation Satellites), Data Cube is a platform for distributing and analyzing resource satellite data by using cloud computing capabilities. Currently a Data Cube has been established in TASA to  manage and operate the data acquired by FORMOSAT series satellites and other open source remote sensing satellites. 

Taiwan Data Cube user interface