Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB

Hyperspectral and Multi-Sensor Remote Sensing

Hyperspectral and Multi-Sensor Remote Sensing

Many remote sensing tasks benefit from simultaneous employment of complementary sensors. An airborne multi-sensor platform acquires considerable image data, being intuitively evaluated by automatic, real-time capable methods.

HYPERSPECTRAL MULTI-SENSOR PLATFORM

Condor airplane

Motor glider Stemme S10 "Condor" and IOSB operator

Many remote sensing tasks, such as environmental monitoring of maritime areas, benefit from increasing the potential information gain by simultaneous employment of complementary sensors. To fulfill these tasks, Fraunhofer IOSB developed a hyperspectral multi-sensor platform (called wingpod) in cooperation with OHB System AG. The wingpod could be easily attached to a Stemme S10 motor glider. Our focus is on designing an easy and intuitive workflow of data acquisition and processing.

 

The wingpod contains an AISA Eagle hyperspectral sensor, a high resolution RGB-camera and an infrared camera combined with IMU and GPS. Capture of these data opens up new opportunities for highly accurate material identification, temperature and position of objects of interest.

Near real-time data processing minimizes the delay between data acquisition, processing and the final information.

 

wingpod wingpod scheme

Multi-sensor platform, called wingpod, und its acquisition scheme

 

 

HYPERSPECTRAL IMAGING

The human eye can distinguish between three different colors: red, green and blue (RGB). A hyperspectral sensor, instead, can differentiate between hundreds of narrow bands of the electro-magnetic spectrum, building up a hyperspectral data cube. This allows a much more accurate distinction between slightly different materials, as well as their physical characteristics.

In the case of environmental monitoring, it is possible to assess the state of health of different vegetation, to map mineral deposits or to detect anomalies, such as localized environmental pollution or floating debris.

 

Fraunhofer IOSB researches and develops already for years automatic evaluation methods for hyperspectral data. This know-ledge consequently influences the development of operational systems. Our focus is on designing an easy and intuitive workflow of data acquisition, pre-processing and evaluation. At the same time research is done on new robust and more efficient evaluation methods, in order to match our customers’ requirements at best.

 

hyperspectral curve hyperspectral cube

Electro-magnetic spectrum with certain visualized channels

Hyperspectral data cube (© OHB System AG)

 

MARITIME SAFETY APPLICATIONS

Flight platforms equipped with the hyperspectral multi-sensor platform offer a cost-effective and permanent monitoring in maritime safety. An important topic thereof is dyke protection. With the help of hyperspectral data processing, the damage on the dyke’s turf cover caused by pest invasion can be detected before it is visible to human eye. Consequently, the necessary precautions can be taken on time.

Another important part of this topic is the monitoring of the soil moisture near the dyke. If the soil is too damp, the firmness of the dyke can be impaired. Regular monitoring against swamp formation can offer an early enough detection of the potential damage to the stability. With the available sensors and its experience in the field, Fraunhofer IOSB is able to perform research on these specific topics.

 oil detection dike control

Oil carpet in a harbor area

Control of vegetation on a dike

FURTHER APPLICATIONS

  • Automatic generation of maps
  • change detection in coastal regions
  • Assessment of environmental pollution
  • vegetation analysis like vegetation stress and detection of algae
  • monitoring of waterways, shallow water and rivers
  • monitoring rivers concerning temperature or fine particles
  • classification and segmentation of minerals
  • detection of artificial objects

 

Talk about analysis of hyperspectral image data on Intergeo 2013
(video in German language only)

 

 

PUBLICATIONS