A high-level fusion method to fuse disparate high-resolution airborne sensor data for change-detection application

A high-level data fusion system that adopts Bayesian statistics involving weights of evidence modelling is described to combine disparate information from airborne digital data such as a digital surface model (DSM), colour (ADS40), thermal infrared (TABI) and hyperspectral images (AISA) of different period. To show the efficacy of the system, an analysis of subtle change-detection is demonstrated. The data fusion system is capable of detecting changes in man-made features automatically in a densely populated area where there are no much prior information. Multi-class segmented images are obtained from the data captured by the four airborne remote sensing sensors. The system performs data fusion modelling by using binary images of each theme classes and thus a total of 40 binary patterns are obtained.

Through Bayesian methods, involving weights of evidence modelling, all the binary images are analysed and finally four binary patterns (indicator images) are identified automatically as significant for the change-detection application. A weighted index overlay model available in the system, combines these four patterns. Data fusion by the weights of evidence modelling is found to be candid and unequivocal for predicting newly transformed locations.

Results of Bayesian method are accurate as the weights are based on statistical analysis. Changes in features such as colour of roof, parking areas, openland areas, newly built structures, and presence or absence of vehicle is automatically extracted by using the high-level data fusion approach. The final predictor image showing the probability of change-detected areas in a high-dense city of Japan is generated robustly.

The fused image products can be used to update maps for digital cartographic applications.

Speaker Details

I received my Ph.D. in 1996 (Remote Sensing) from the Indian Institute of Technology-Bombay, Mumbai, India. My research interests include Digital Image Processing, Computer vision, 3D modelling, datafusion, GIS and developing software systems for high-resolution digital remote sensing data processing.I have received Japanese Monbusho scholarship and JSPS fellowship to carryout two strong post-doctoral research projects in Remote Sensing and GIS at the GIS Laboratory of Keio University in Japan. I chaired, Medical Imaging session of Computer Graphics and Imaging (CGIM 2000) conference and International Symposium on Digital Earth (ISDE-J). I am a Peer Reviewer for the International Journal of Remote Sensing (IJRS), UK, Pattern Analysis and Applications Journal (UK) and SCI (Systemics, Cybernetics and Informatics) -USA conferences. I am also the Regional Correspondent (Japan) for Geomatics International Magazine- GIM (The Netherlands).I am a member of Japan Society for Promotion of Science (JSPS), Japan Society for Photogrammetry and Remote Sensing (JSPRS) and, Association for Real-time Imaging and Dynamic Analysis- ARIDA.Currently I am a Project Manager with the GIS Institute of PASCO Corporation in Tokyo.

Date:
Speakers:
Babu Madhavan
Affiliation:
GIS Institute of PASCO Corporation
    • Portrait of Jeff Running

      Jeff Running