Paper
FNEA and D-S Based Remote Sensing Approach of Land-use Classification
- Authors:
- Haiqing He; Shengxiang Huang
- Abstract
- Extracted information by remote sensing classification has become a widespread approach for land-use. However, remote sensing images have mixed pixels, same object with different spectrums and different objects with same spectrum, which may weaken remote sensing classification accuracy. In this paper a new method of remote sensing land-use classification was proposed based on fractal net evolution approach and Dempster Shafer evidence theory. Remote sensing image was segmented to sub-objects in the method of fractal net evolution approach, and a frame of discernment was established according of the Land-use Classification. Evidence, the core of the frame, was extracted from the data’s special and spectral characters of sub-objects in multi-source. Basic probability assignment was calculated to establish a decision-making rule, which was used to classify the sub-objects. An example of classifying land-use with the satellite data from SPOT-5 and RADARSAT-2 was carried out. Test result shows that the method improved the accuracy of RS image classification with higher application value.
- Keywords
- FNEA (Fractal Net Evolution Approach); Dempster Shafer Evidence Theory; BPA (Basic Probability Assignment); Land Classification
- StartPage
- 54
- EndPage
- 60
- Doi