News |  Sitemap |  Contact
PDF Export  | 

Knut Kvaal: A plugin system for (1-D, 2-D) texture and signal characterisation using AMT (Angle Measure Technique)


Texture characterisation plays an important role in image analysis as do signal complexity characterisation for 1-dimensional series (time series a.o.). One important data analysis task concerns comparison of different characterisation algorithms, f.ex. optimal discrimination between textures and texture classes. Texture algorithms are often specifically applied to international benchmark data such as the well-known Brodatz album etc. Other types of image are often constructed for more narrow purposes, e.g. to measure one specific feature extractor. An ImageJ plugin for AMT has been designed and implemented containing all algorithm and application options suggested in the 10 year history of the Angle Measure Technique. The AMT plugin will be used in conjunction with Graylevel Coocurence Matrix (GLCM) plugin as a new proposal in the AMT research. The output from the AMT plugin is further processed in PLS Discriminant analysis and Linear Discriminant Analysis (LDA) to perform a classification of the dataset in scope. The plugins are a user friendly, graphically oriented visual data analysis. In this paper we describe the implementation concepts and give a short introduction to the potential use of the plugin. In our work we will present both benchmark and practical use of the plugin implementation.


Texture characterisation, Angle Measure Technique (AMT), GLCM, PLS, LDA

Administrative data

Presenting author: Knut Kvaal
Organisation: Norwegian University of Life Sciences (UMB)

co-authors: Andreas Floe

Type: Poster (portrait)

© Luxembourg Institute of Science and Technology | Legal Notice