An improved method in change detection of multitemporal remote sensing image

Fangshun Liao, Sufen Yu, Ying Li, Yanning Zhang

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

Abstract

Traditional Markov random Field (MRF) methods assume that neighboring pixels tend to have the same label. However, this assumption is always inconsistent with the actual situation and affects the resultant accuracy of the algorithm. To overcome this, we propose an object-based Markov Random Field (OMRF) model and a change detection method based on OMRF model. The OMRF model assumes that pixels within same object are in the same class. First, we generate the difference image from multi-temporal remote sensing images. Second, Mean Shift is applied to extract objects from difference image. Finally, change detection map is generated by iterative algorithm. The experimental results show that the algorithm can effectively improve the detection accuracy of the algorithm on real remote sensing datasets.

Original languageEnglish
Title of host publicationIntelligence Science and Big Data Engineering - 4th International Conference, IScIDE 2013, Revised Selected Papers
PublisherSpringer Verlag
Pages587-594
Number of pages8
ISBN (Print)9783642420566
DOIs
StatePublished - 2013
Event4th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2013 - Beijing, China
Duration: 31 Jul 20132 Aug 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8261 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference4th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2013
Country/TerritoryChina
CityBeijing
Period31/07/132/08/13

Keywords

  • Change Detection
  • Markov Random Field
  • Object-based
  • Remote Sensing Images

Fingerprint

Dive into the research topics of 'An improved method in change detection of multitemporal remote sensing image'. Together they form a unique fingerprint.

Cite this