FEATURE SELECTION IN CT IMAGES BASED ON PARTICLE SWARM BASED OPTIMIZATION
Journal Title: Elysium Journal of Engineering Research and Management - Year 2016, Vol 3, Issue 3
Abstract
Lung diseases are the common disorder that can affect the lungs and create breathing problem in all the humans. CT images plays an important role for diagnosing the lung diseases.This project proposes a Particle swarm based optimization technique to select the best features. Initially the features can be extracted based on B-HOG features, Wavelet features, LBP features and CVH features. The feature selection process were employed based on is Particle swarm based optimization (PSO) is the objective function. The selected features were then classified using different classifiers like SVM, Artificial neural network and Fuzzy nearest neighbourhood. For all the considered classifiers, our PSO method brought the better recognition. The advantages on computation effectiveness and efficiency of PSO are shown through experiments. The performance analysis is to calculate the accuracy, sensitivity and specificity.
Authors and Affiliations
Cecil Theijas M. , Varatharajan R.
Keywords
Related Articles
- EP ID EP365607
- DOI -
- Views 123
- Downloads 0
Elimination of Commutation Failure of VSC HVDC System with Controllable Capacitor
This system presents a totally unique hybrid converter configuration for traditional Line-Commutated converter (LCC) HVDC technology aiming to eliminate commutation failures at lower place serious faults. Dynamic series...
AN EFFICIENT WI-FI INTERNET ACCESS FOR MOVING VEHICLES USING SWIMMING SCHEME
Accessing internet from moving vehicles is demanding and has been rapidly growing. Due to mobile detonation, the overloading issue of cellular networks is increasing. So, the Wi-Fi based accessing is co...
XML DATA PROCESSING APPROACH FOR AUTOMATIC IDENTIFICATION OF CARDIAC ABNORMALITIES IN ECG
In medical applications, the process that records the electrical activity of the heart is termed as Electrocardiography (ECG). The traditional way of ECG is in the form of the transthoracic interpretatio...
DETECTION OF BLEEDING USING MORPHOLOGICAL AND SEGMENTATION IN WCE VIDEO
This paper describes propose a method to detect bleeding regions from WCE video. To find bleeding regions we are using super pixel segmentation process. Image segmentation is the process of dividing...
EFFECTIVE LEARNING OF RESTRAINTS FOR SEMI ADMINISTERED CLUSTERING
Semi-supervised clustering is improving clustering performance by considering user supervision in the form of pairwise constraints. We study the active learning problem of selecting pairwise must-link and cannot...