Clustering and classifying images with local and global variability

A procedure for clustering and classifying images determined by three classification variables is presented. A measure of global variability based on the singular value decomposition of the image matrices, and two average measures of local variability based on spatial correlation and spatial changes. The performance of the procedure is compared using three different databases.

MoreLess

Year of publication:	2009-01
Authors:	Giuliodori, Andrea ; Lillo, Rosa E. ; Peña, Daniel
Institutions:	Departamento de Estadistica, Universidad Carlos III de Madrid
Subject:	Images \| Cluster \| Classification

More details

Extent:	application/pdf
Series:	Statistics and Econometrics Working Papers.
Type of publication:	Book / Working Paper
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10005196578