Testing the performance of image representations for mass classification in digital mammograms

Angelini, Enrico ; Campanini, Renato ; Iampieri, Emiro ; Lanconelli, Nico ; Masotti, Matteo ; Roffilli, Matteo (2004) Testing the performance of image representations for mass classification in digital mammograms. [Preprint]
Warning

There is another version of this document. Click here to view it.

Full text available as:
[img]
Preview
PDF
Download (219kB) | Preview

Abstract

In this paper a two-class classification problem is faced. One class is constituted by tumoral masses, breast tumors with size ranging from 3 mm to 30 mm. The other class is constituted by non-masses. A Support Vector Machine (SVM) is used as a classifier. Both, masses and non-masses, are extracted from the University of South Florida (USF) mammographic image database and are presented to the classifier as crops with pixel size 64 x 64. In order to find the optimal solution to this problem, different featureless crops representations are evaluated. In particular, a pixel-based representation, a Discrete Wavelet Transform (DWT) representation and an Overcomplete Wavelet Transform (OWT) representation are tested.

Abstract
Document type
Preprint
Creators
CreatorsAffiliationORCID
Angelini, Enrico
Campanini, Renato
Iampieri, Emiro
Lanconelli, Nico
Masotti, Matteo
Roffilli, Matteo
Keywords
Wavelets, Image Processing, Support Vector Machine, Pattern Classification, Computer-Aided Detection in Mammography
Subjects
DOI
Deposit date
21 Oct 2004
Last modified
16 May 2011 11:37
URI

Other metadata

Available versions of this document

Downloads

Downloads

Staff only: View the document

^