Change search
ReferencesLink to record
Permanent link

Direct link
Estimation of polygons and areas
Global Biometric Sciences, Bristol-Myers Squibb Company.
Umeå University, Faculty of Science and Technology, Department of Mathematics and Mathematical Statistics.
Umeå University, Faculty of Science and Technology, Department of Computing Science.ORCID iD: 0000-0002-7657-6917
2006 (English)In: Scandinavian Journal of Statistics, ISSN 0303-6898, E-ISSN 1467-9469, Vol. 33, no 3, 541-559 p.Article in journal (Refereed) Published
Abstract [en]

In forestry the problem of estimating areas is central. This paper addresses area estimation through fitting of a polygon to observed coordinate data. Coordinates of corners and points along the sides of a simple closed polygon are measured with independent random errors. This paper focuses on procedures to adjust the coordinates for estimation of the polygon and its area. Different new techniques that consider different amounts of prior information are described and compared. The different techniques use restricted least squares, maximum likelihood and the expectation maximization algorithm. In a simulation study it is shown that the root mean square errors of the estimates are decreased when coordinates are adjusted before estimation. Minor further improvement is achieved by using prior information about the order and the distribution of the points along the sides of the polygon. This paper has its origin in forestry but there are also other applications.

Place, publisher, year, edition, pages
John Wiley & Sons, 2006. Vol. 33, no 3, 541-559 p.
Keyword [en]
area estimation, dirichlet distribution, expectation maximization algorithm, iteratively reweighted least squares, order restricted inference, polygon measurements, quadratic constraint, restricted least-square methods
National Category
Probability Theory and Statistics
URN: urn:nbn:se:umu:diva-21968DOI: 10.1111/j.1467-9469.2006.00507.xISBN: 0303-6898OAI: diva2:212227
Available from: 2009-04-21 Created: 2009-04-21 Last updated: 2014-06-11Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full text

Search in DiVA

By author/editor
Bondesson, LennartBörlin, Niclas
By organisation
Department of Mathematics and Mathematical StatisticsDepartment of Computing Science
In the same journal
Scandinavian Journal of Statistics
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 55 hits
ReferencesLink to record
Permanent link

Direct link