A texture-based approach for extracting residential areas from high-resolution imagery

Gu Juan*, Chen Jun, Zhang Hongwei, Qiming ZHOU

*Corresponding author for this work

Research output: Chapter in book/report/conference proceedingConference contributionpeer-review

Abstract

When spectral classification approach has lost its efficiency in object recognition from finer resolution imagery, texture plays more and more important role. This paper introduces an approach of using the combination of four kinds of texture measurements (grey mean value, standard deviation, edge point density, edge line density) to extracting residential areas from high-resolution imagery. The job of extraction is performed in three steps: (1) Extraction of candidate pixels of residential areas based on mean grey value and standard deviation; (2) Extraction of candidate areas of residential areas based on edge point density; (3) Verification of candidate areas of residential areas based on edge line density. An experiment by using IKONOS PAN image is given to verify the correct of the method.

Original languageEnglish
Title of host publicationGeoinformatics 2006
Subtitle of host publicationRemotely Sensed Data and Information
DOIs
Publication statusPublished - 2006
EventGeoinformatics 2006: Remotely Sensed Data and Information - Wuhan, China
Duration: 28 Oct 200629 Oct 2006

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume6419
ISSN (Print)0277-786X

Conference

ConferenceGeoinformatics 2006: Remotely Sensed Data and Information
Country/TerritoryChina
CityWuhan
Period28/10/0629/10/06

Scopus Subject Areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

User-Defined Keywords

  • Edge line density
  • Edge point density
  • Grey mean value
  • Residential areas
  • Stand deviation
  • Textures

Fingerprint

Dive into the research topics of 'A texture-based approach for extracting residential areas from high-resolution imagery'. Together they form a unique fingerprint.

Cite this