Skip to main navigation Skip to search Skip to main content

Improved two-dimensional Otsu based on firefly optimization for low signal-to-noise ratio images

  • Harbin Institute of Technology Weihai
  • Jiangsu Automation Research Institute

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

To improve two-dimensional (2D) Otsu thresholding’s performance in both computation speed and segmentation quality, an improved 2D Otsu algorithm is proposed for low Signal-to-noise Ratio (SNR) images. A new 2D histogram is defined based on median gray-scale and Gaussian average gray-scale. By meeting better to the assumption of that the object’s probability and the background’s probability sum up to 1, the new 2D histogram enhances the thresholding algorithm’s robustness to severe noise. Then a scheme of calculating the fitness function based on firefly optimization algorithm is employed to search for optimal thresholds. The proposed algorithm is applied to typical low SNR images–microscopic images of ocean plankton, and to Lenna test image. Experiment results show that with better thresholding quality, the running time of the proposed algorithm is reduced to 2.5% of the conventional 2D Otsu.

Original languageEnglish
Title of host publicationAdvances in Swarm Intelligence - 8th International Conference, ICSI 2017, Proceedings
EditorsYing Tan, Hideyuki Takagi, Yuhui Shi
PublisherSpringer Verlag
Pages611-617
Number of pages7
ISBN (Print)9783319618234
DOIs
StatePublished - 2017
Event8th International Conference on Swarm Intelligence, ICSI 2017 - Fukuoka, Japan
Duration: 27 Jul 20171 Aug 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10385 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference8th International Conference on Swarm Intelligence, ICSI 2017
Country/TerritoryJapan
CityFukuoka
Period27/07/171/08/17

Keywords

  • Firefly optimization
  • Plankton microscopic image analysis
  • Two dimensional Otsu thresholding

Fingerprint

Dive into the research topics of 'Improved two-dimensional Otsu based on firefly optimization for low signal-to-noise ratio images'. Together they form a unique fingerprint.

Cite this