|
|
|
|
|
|
|
RDA is crucial for many applications in a collaborative and distributed sensor network consisting of UAVs and fixed/static sensors. We presents a new approach to data fusion for automatic recognition, surveillance, and tracking research areas in intelligent transportation systems. Robust data alignment (RDA), finding relational maps among a sequence of invariant feature data sets, is one of the key requirements for successful data fusion. To achieve RDA for correspondence-less data fusion, we construct a cost criterion based on information theory and solve an optimization problem with a cooperative search strategy. Experimental results on a video sequence collected from an unmanned aerial vehicle (UAV) indicate the potential of aerial monitoring and tracking systems built upon our information-theoretic RDA.
The figures below illustrate the following:
Video registration with weighted features
Multimodal data fusion: Multimodality between Electro Optical sensor and Infrared sensor has been investigated. The figure illustrates the registration between the two different modalities.
Sangil Jwa, Ümit Özgüner, and Zhijun Tang, "Information-theoretic Data Registration for UAV-based sensing," IEEE Trans. on Intelligent Transportation Systems, Vol. 9, No. 1, March 2008, pp. 5-15.
Sangil Jwa and Ümit Özgüner, "Multi-UAV Sensing Over Urban Areas Via Layered Data Fusion," IEEE Statistical Signal Processing Workshop, August 26-29, Madison, WI, 2007, pp. 576-580.
Sangil Jwa, Zhijun Tang, and Ümit Özgüner, "Robust Data Alignment Based on Information Theory and Its Applications in Road Following Situation," IEEE International Conference on Intelligent Transportation Systems, Toronto, Canada, September 17-20, 2006, pp. 1328-1333.