Name
Abstract | ||
|---|---|---|
The problem of off-grid target detection with the normalized matched filter (NMF) detector is considered. We show that this detector is highly sensitive to off-grid targets. In particular its mean asymptotic detection probability may not converge to 1. We then consider two solutions to solve this off-grid problem. The first solution approximates the Generalized Likelihood Ratio Test (GLRT) by oversampling the resolution cell; this solution may be computationally heavy and does not permit to compute a theoretical detection threshold. We then propose a second solution based on the matched subspace detection framework. For Doppler steering vectors, the subspace considered is deduced from Discrete Prolate Spheroidal Sequence vectors. Simulation results permit to demonstrate interesting performance for off-grid targets. |
Year | Venue | Keywords |
|---|---|---|
2016 | European Signal Processing Conference | Off-grid,normalized matched filter,matched subspace detector,discrete prolate spheroidal sequences |
Field | DocType | ISSN |
Object detection,Likelihood-ratio test,Subspace topology,Pattern recognition,Oversampling,Algorithm,Non-negative matrix factorization,Artificial intelligence,Matched filter,Detector,Gaussian noise,Mathematics | Conference | 2076-1465 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Olivier Rabaste | 1 | 6 | 3.50 |
| Jonathan Bosse | 2 | 17 | 5.77 |
| Jean Philippe Ovarlez | 3 | 190 | 25.11 |