Automatic Radar Target Recognition from Predicted HRR Range Profiles
In this project we research new methods for the use of high-resolution radar range profiles for radar target recognition. Our research focuses on the use of radar data predicted from CAD models and electro-magnetic simulation software as prototype data for the classification of measured data. Current research topics include model simplification for real-time target recognition, and statistical analysis of the difference between measured and predicted data.
Research group members
drs. J. Portegies Zwart
dr. ir. B. Kröse
prof. dr. ir. F.C.A. Groen
This project is funded by TNO Physics and Electronics Laboratory.
We described a new way of estimating mixture model parameters when the relationshiop between latent variable and feature vector is known a priori. This work was presented at the ICPR 2000 Conference in september 2000 in Barcelona, Spain.
Joris Portegies Zwart and Ben Kröse.
Constrained Mixture Modeling of Intrinsically Low-Dimensional Distributions.
15th International Conference on Pattern Recognition, Barcelona, Spain. September 2000.
Joris Portegies Zwart homepage