Modified fuzzy c-means classification technique for mapping vague wetlands using Landsat ETM+ imagery
✍ Scribed by Wen-Ya Chiu; Isabelle Couloigner
- Publisher
- John Wiley and Sons
- Year
- 2006
- Tongue
- English
- Weight
- 770 KB
- Volume
- 20
- Category
- Article
- ISSN
- 0885-6087
- DOI
- 10.1002/hyp.6378
No coin nor oath required. For personal study only.
✦ Synopsis
Abstract
Wetland mapping derived from remotely sensed images is subject to error and uncertainty. Fuzzy classification techniques can deal with the spectral and spatial vagueness, and can be used to model the uncertainty in remote sensing classification. In this paper, we present a modified Fuzzy C‐Means (FCM) classifier that allows the local texture information to regularize the classification result iteratively, and we propose a threshold defuzzification method to extract the potential wetland areas from Landsat‐7 Enhanced Thematic Mapper Plus (ETM+) imagery. By introducing the transition classes during the defuzzification process, the modified classifier reduces commission‐errors and improves the mapping accuracy when compared to the standard FCM classifier. The accuracy assessment and a test of statistical significance indicate that the modified FCM classifier shows a significantly better mapping result than the standard FCM classifier because of the incorporation of this spatial information. Copyright © 2006 John Wiley & Sons, Ltd.