Content Based Image Retrieval for Lung CT Slices using Texture Features

Abstract

Content Based Image Retrieval (CBIR) is used to retrieve the relevant CT slices from the database based on the query CT slice. In the proposed method, the database consists of diseased CT slices of Cavitary TB. The processes involved in the proposed method include preprocessing, segmentation, ROI extraction and slice retrieval. The lung CT slice is segmented using adaptive thresholding and the diseased regions are extracted from the segmented CT slice by Harris corner point detection. The combination of Block Difference of Inverse Probabilities (BDIP) and Tamura feature extraction methods are proposed for the extracted diseased region and the segmented lung. The Euclidean distance measures the similarity between the features of the query CT slice and the database CT slices. Depending on the distance measures, CT slices are retrieved from the database. The performance of the proposed feature extraction methods are calculated using precision and recall parameters for better retrieval of CT slices.

Authors and Affiliations

Anita Titus, M. Bruntha

Keywords

Related Articles

Adaptive Simplified Successive Cancellation for Polar Codes Based on Frozen Bits

Polar codes have been recently proposed as the first low complexity of error correcting codes for wireless channels that can provably achieve the capacity of symmetric binary-input memory less channels. Here, the bit er...

slugSurgically Altered Face Image Detection U sing Genetic Algorithm– A Comprehensive Study

In recent years, plastic surgery has become popular worldwide. People take facial plastic surgery to correct feature defects or improve attractiveness and confidence. It has been observed that many face...

DC-DC Converter Based SHE-PWM Technique Multilevel Inverter

This paper presents a Selective Harmonic Elimination-PWM technology (SHE-PWM) employing H-Bridge multilevel inverter configuration. In this method variable DC voltage is used and the lower order harmonics are eliminated...

A Trusted Framework for Authentication and Security for Business Applications in Cloud

Cloud stands for Common Location independent Online Utility on Demand services (or) Communities Libraries Online Union Database which is abbreviated from AT&T (American Telephone & Telegraph) corporation. Now a day’s cl...

Analysis on Security Evaluation of Pattern Classifiers under Attack

Analysis on security evaluation of pattern classifiers under attack describes pattern classification systems that are security evaluation problems because of different attacks. Pattern Classification commonly used in ad...

Download PDF file
  • EP ID EP21016
  • DOI -
  • Views 256
  • Downloads 5

How To Cite

Anita Titus, M. Bruntha (2015). Content Based Image Retrieval for Lung CT Slices using Texture Features. International Journal for Research in Applied Science and Engineering Technology (IJRASET), 3(6), -. https://europub.co.uk/articles/-A-21016