Microscopic image classification via WT-based covariance descriptors using Kullback-Leibler distance
Author
Keskin, Furkan
Çetin, A. Enis
Erşahin, Tülin
Çetin-Atalay, Rengül
Date
2012Source Title
ISCAS 2012 - 2012 IEEE International Symposium on Circuits and Systems
Publisher
IEEE
Pages
2079 - 2082
Language
English
Type
Conference PaperItem Usage Stats
133
views
views
97
downloads
downloads
Abstract
In this paper, we present a novel method for classification of cancer cell line images using complex wavelet-based region covariance matrix descriptors. Microscopic images containing irregular carcinoma cell patterns are represented by randomly selected subwindows which possibly correspond to foreground pixels. For each subwindow, a new region descriptor utilizing the dual-tree complex wavelet transform coefficients as pixel features is computed. WT as a feature extraction tool is preferred primarily because of its ability to characterize singularities at multiple orientations, which often arise in carcinoma cell lines, and approximate shift invariance property. We propose new dissimilarity measures between covariance matrices based on Kullback-Leibler (KL) divergence and L 2-norm, which turn out to be as successful as the classical KL divergence, but with much less computational complexity. Experimental results demonstrate the effectiveness of the proposed image classification framework. The proposed algorithm outperforms the recently published eigenvalue-based Bayesian classification method. © 2012 IEEE.
Keywords
Bayesian classificationCancer cell lines
Carcinoma cell lines
Carcinoma cells
Classification framework
Covariance matrices
Descriptors
Dissimilarity measures
Dual-tree complex wavelet transform
KL-divergence
Kullback-Leibler distance
Kullback-Leibler divergence
Microscopic image
Multiple orientations
Region covariance matrixes
Shift invariance
Cell culture
Covariance matrix
Eigenvalues and eigenfunctions
Feature extraction
Image classification
Pixels