Name
Abstract | ||
|---|---|---|
In previous work, the statistical characteristics of the background or the noise under H0 hypothesis are similar as that under H1 hypothesis. Accordingly, the parameters under both hypotheses are estimated by the maximum likelihood method and finally a generalized likelihood ratio test based detector is developed, such as the matched subspace detector. Unfortunately, this kind of statistical similarity for both hypotheses may be changing, which is directly related to the unknown beforehand target fill factor. A hypothesis independent method is proposed to solve this problem, which uses different approaches to estimate the parameters for different hypotheses. Experiments on simulated data and real hyperspectral image demonstrate the ability of this proposed detector for subpixel target detection. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1016/j.sigpro.2014.08.018 | Signal Processing |
Keywords | Field | DocType |
Hyperspectral image,Hypothesis independent,Maximum likelihood method,Matched subspace detector,Subpixel target detection | Subspace topology,Likelihood-ratio test,Pattern recognition,Variance estimation,Maximum likelihood,Hyperspectral imaging,Artificial intelligence,Subpixel rendering,Detector,Mathematics | Journal |
Volume | Issue | ISSN |
110 | C | 0165-1684 |
Citations | PageRank | References |
13 | 0.56 | 18 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Bo Du | 1 | 1662 | 130.01 |
| Yuxiang Zhang | 2 | 167 | 15.28 |
| Liangpei Zhang | 3 | 5448 | 307.02 |
| Lefei Zhang | 4 | 840 | 47.83 |