Log-ratio descriptor for tissues and organs images

Research output: Contribution to journalArticle

Abstract

A unique feature descriptor is very important to ensure successful matching of the points of interest. Current approaches of feature descriptor that depend on intensity information of the grayscale image are not able to differentiate the feature for tissues and organs images (TOI). The main reason is TOI have homogeneous property of texture, surface and color. Therefore, in this paper a novel descriptor for tissues and internal organs matching, namely set vector log-ratio descriptor is proposed. The descriptor focuses on the enhancement of the intensity feature through a fusion of log normal distribution and ratio proportion approach. Experimental results show that the proposed descriptor achieves more than 90% recognition rate for all datasets and average error distance as low as 4.34 pixels. Besides, the validation results also confirmed that the algorithm performs relatively well compared to existing descriptors.

Original languageEnglish
Pages (from-to)3912-3916
Number of pages5
JournalAdvanced Science Letters
Volume23
Issue number5
DOIs
Publication statusPublished - 1 May 2017

Fingerprint

Descriptors
Tissue
Surface Properties
Normal Distribution
Normal distribution
pixel
Fusion reactions
Color
Textures
texture
Pixels
Surface Texture
Log Normal Distribution
organ
tissue
Differentiate
Fusion
Proportion
Enhancement
Pixel

Keywords

  • Descriptor
  • Feature matching
  • Log normal distribution
  • Log-ratio
  • Organ image
  • Tissue image

ASJC Scopus subject areas

  • Health(social science)
  • Computer Science(all)
  • Education
  • Mathematics(all)
  • Environmental Science(all)
  • Engineering(all)
  • Energy(all)

Cite this

Log-ratio descriptor for tissues and organs images. / Karim, Rohana Abdul; Mustafa, Mohd. Marzuki; Zulkifley, Mohd Asyraf.

In: Advanced Science Letters, Vol. 23, No. 5, 01.05.2017, p. 3912-3916.

Research output: Contribution to journalArticle

@article{4ea431aef8f54ef0a6181a4fff0bbf23,
title = "Log-ratio descriptor for tissues and organs images",
abstract = "A unique feature descriptor is very important to ensure successful matching of the points of interest. Current approaches of feature descriptor that depend on intensity information of the grayscale image are not able to differentiate the feature for tissues and organs images (TOI). The main reason is TOI have homogeneous property of texture, surface and color. Therefore, in this paper a novel descriptor for tissues and internal organs matching, namely set vector log-ratio descriptor is proposed. The descriptor focuses on the enhancement of the intensity feature through a fusion of log normal distribution and ratio proportion approach. Experimental results show that the proposed descriptor achieves more than 90{\%} recognition rate for all datasets and average error distance as low as 4.34 pixels. Besides, the validation results also confirmed that the algorithm performs relatively well compared to existing descriptors.",
keywords = "Descriptor, Feature matching, Log normal distribution, Log-ratio, Organ image, Tissue image",
author = "Karim, {Rohana Abdul} and Mustafa, {Mohd. Marzuki} and Zulkifley, {Mohd Asyraf}",
year = "2017",
month = "5",
day = "1",
doi = "10.1166/asl.2017.8257",
language = "English",
volume = "23",
pages = "3912--3916",
journal = "Advanced Science Letters",
issn = "1936-6612",
publisher = "American Scientific Publishers",
number = "5",

}

TY - JOUR

T1 - Log-ratio descriptor for tissues and organs images

AU - Karim, Rohana Abdul

AU - Mustafa, Mohd. Marzuki

AU - Zulkifley, Mohd Asyraf

PY - 2017/5/1

Y1 - 2017/5/1

N2 - A unique feature descriptor is very important to ensure successful matching of the points of interest. Current approaches of feature descriptor that depend on intensity information of the grayscale image are not able to differentiate the feature for tissues and organs images (TOI). The main reason is TOI have homogeneous property of texture, surface and color. Therefore, in this paper a novel descriptor for tissues and internal organs matching, namely set vector log-ratio descriptor is proposed. The descriptor focuses on the enhancement of the intensity feature through a fusion of log normal distribution and ratio proportion approach. Experimental results show that the proposed descriptor achieves more than 90% recognition rate for all datasets and average error distance as low as 4.34 pixels. Besides, the validation results also confirmed that the algorithm performs relatively well compared to existing descriptors.

AB - A unique feature descriptor is very important to ensure successful matching of the points of interest. Current approaches of feature descriptor that depend on intensity information of the grayscale image are not able to differentiate the feature for tissues and organs images (TOI). The main reason is TOI have homogeneous property of texture, surface and color. Therefore, in this paper a novel descriptor for tissues and internal organs matching, namely set vector log-ratio descriptor is proposed. The descriptor focuses on the enhancement of the intensity feature through a fusion of log normal distribution and ratio proportion approach. Experimental results show that the proposed descriptor achieves more than 90% recognition rate for all datasets and average error distance as low as 4.34 pixels. Besides, the validation results also confirmed that the algorithm performs relatively well compared to existing descriptors.

KW - Descriptor

KW - Feature matching

KW - Log normal distribution

KW - Log-ratio

KW - Organ image

KW - Tissue image

UR - http://www.scopus.com/inward/record.url?scp=85023751496&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85023751496&partnerID=8YFLogxK

U2 - 10.1166/asl.2017.8257

DO - 10.1166/asl.2017.8257

M3 - Article

VL - 23

SP - 3912

EP - 3916

JO - Advanced Science Letters

JF - Advanced Science Letters

SN - 1936-6612

IS - 5

ER -