Détail de l'auteur
Auteur T.G. Van Niel |
Documents disponibles écrits par cet auteur



On the relationship between training sample size data dimensionality: Monte Carlo analysis of broadland multi-temporal classification / T.G. Van Niel in Remote sensing of environment, vol 98 n° 4 (30/10/2005)
[article]
Titre : On the relationship between training sample size data dimensionality: Monte Carlo analysis of broadland multi-temporal classification Type de document : Article/Communication Auteurs : T.G. Van Niel, Auteur ; T.R. Mcvicar, Auteur Année de publication : 2005 Article en page(s) : pp 468 - 480 Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Traitement d'image optique
[Termes descripteurs IGN] analyse diachronique
[Termes descripteurs IGN] classification par maximum de vraisemblance
[Termes descripteurs IGN] échantillonnage d'image
[Termes descripteurs IGN] figure géométrique
[Termes descripteurs IGN] image Landsat-ETM+
[Termes descripteurs IGN] image multitemporelle
[Termes descripteurs IGN] méthode de Monte-CarloRésumé : (Auteur) The number of training samples per class (n) required for accurate Maximum Likelihood (ML) classification is known to be affected by the number of bands (p) in the input image. However, the general rule which defines that n should be 10p to 30p is often enforced universally in remote sensing without questioning its relevance to the complexity of the specific discrimination problem. Furthermore, identifying this many training samples is often problematic when many classes and/or many bands are used. It is important, then, to test how this generally accepted rule matches common remote sensing discrimination problems because it could be unnecessarily restrictive for many applications. This study was primarily conducted in order to test whether the general rule defining the relationship between n and p was well-suited for ML classification of a relatively simple remote sensing-based discrimination problem. To summarise the mean response of n-to-p for our study site, a Monte Carlo procedure was used to randomly stack various numbers of bands into thousands of separate image combinations that were then classified using an ML algorithm. The bands were randomly selected from a 119-band Enhanced Thematic Mapper-plus (ETM+) dataset comprised of 17 images acquired during the 2001-2002 southern hemisphere summer agricultural growing season over an irrigation area in south-eastern Australia. Results showed that the number of training samples needed for accurate ML classification was much lower than the current widely accepted rule. Due to the asymptotic nature of the relationship, we found that 95% of the accuracy attained using n = 30p samples could be achieved by using approximately 2p to 4p samples, or Numéro de notice : A2005-434 Thématique : IMAGERIE Nature : Article nature-HAL : ArtAvecCL-RevueIntern Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=27570
in Remote sensing of environment > vol 98 n° 4 (30/10/2005) . - pp 468 - 480[article]Réservation
Réserver ce documentExemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité 110-05181 RAB Revue Centre de documentation En réserve 3L Disponible Experimental evaluation of positional accuracy estimates from linear network using point- and line-based testing methods / T.G. Van Niel in International journal of geographical information science IJGIS, vol 16 n° 5 (july 2002)
[article]
Titre : Experimental evaluation of positional accuracy estimates from linear network using point- and line-based testing methods Type de document : Article/Communication Auteurs : T.G. Van Niel, Auteur ; T.R. Mc Vicar, Auteur Année de publication : 2002 Article en page(s) : pp 455 - 473 Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Information géographique
[Termes descripteurs IGN] analyse de données
[Termes descripteurs IGN] estimation de position
[Termes descripteurs IGN] objet géographique linéaire
[Termes descripteurs IGN] point d'appui
[Termes descripteurs IGN] précision géométrique (imagerie)
[Termes descripteurs IGN] primitive géométrique
[Termes descripteurs IGN] propagation d'erreur
[Termes descripteurs IGN] test statistiqueRésumé : (Auteur) Positional accuracy estimates of linear features based on their welldefined points can be significantly different than the accuracy estimates determined from their adjoining lines. A case study was conducted to determine both the cause of this difference as well as the relative effect of the number of points used in calculating the point-based accuracy estimate. Results showed that : (1) the difference between accuracy estimates was primarily due to dissimilar criteria for measuring closeness of primitive features ; and (2) the current US requirement of using 20 well-defined points provided a reasonable estimation of accuracy for the case study. The differences in accuracy estimates described in this study should impact both the definition of geospatial accuracy standards, and the interpretation of geographical analyses with respect to error propagation. Numéro de notice : A2002-149 Thématique : GEOMATIQUE Nature : Article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=22064
in International journal of geographical information science IJGIS > vol 16 n° 5 (july 2002) . - pp 455 - 473[article]Réservation
Réserver ce documentExemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité 079-02051 RAB Revue Centre de documentation En réserve 3L Disponible