| タイトル | The analytical design of spectral measurements for multispectral remote sensor systems |
| 本文（外部サイト） | http://hdl.handle.net/2060/19790017278 |
| 著者（英） | Landgrebe, D. A.; Wiersma, D. J. |
| 著者所属（英） | Purdue Univ. |
| 発行日 | 1979-01-01 |
| 言語 | eng |
| 内容記述 | The author has identified the following significant results. In order to choose a design which will be optimal for the largest class of remote sensing problems, a method was developed which attempted to represent the spectral response function from a scene as accurately as possible. The performance of the overall recognition system was studied relative to the accuracy of the spectral representation. The spectral representation was only one of a set of five interrelated parameter categories which also included the spatial representation parameter, the signal to noise ratio, ancillary data, and information classes. The spectral response functions observed from a stratum were modeled as a stochastic process with a Gaussian probability measure. The criterion for spectral representation was defined by the minimum expected mean-square error. |
| NASA分類 | EARTH RESOURCES AND REMOTE SENSING |
| レポートNO | 79N25449
E79-10200
TR-EE-79-13
NASA-CR-160170
LARS-TR-122678 |
| 権利 | No Copyright |
