A neural approach to image thresholding

Ahmed A. Othman, Hamid R. Tizhoosh

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Image thresholding (as the simplest form of segmentation) is a very challenging task because of the differences in the characteristics of different images such that different thresholds may be tried to obtain maximum segmentation accuracy. In this paper, a supervised neural network is used to "dynamically" threshold images by assigning a suitable threshold to each image. The network is trained using a set of simple features extracted from medical images randomly selected form a sample set and then tested using the remaining medical images. The results are compared with the Otsu algorithm and the active shape models (ASM) approach.

Original languageEnglish (US)
Title of host publicationArtificial Neural Networks, ICANN 2010 - 20th International Conference, Proceedings
Number of pages4
EditionPART 1
StatePublished - 2010
Event20th International Conference on Artificial Neural Networks, ICANN 2010 - Thessaloniki, Greece
Duration: Sep 15 2010Sep 18 2010

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 1
Volume6352 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference20th International Conference on Artificial Neural Networks, ICANN 2010

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'A neural approach to image thresholding'. Together they form a unique fingerprint.

Cite this