[1]GUO Peng,XU Li-ping,WANG Lin,et al.Research for Object-Oriented TM Image Segmentation and Information Extraction Technology[J].Research of Soil and Water Conservation,2013,20(03):120-124.
Copy
Research of Soil and Water Conservation[ISSN 1005-3409/CN 61-1272/P] Volume:
20
Number of periods:
2013 03
Page number:
120-124
Column:
Public date:
2013-06-28
- Title:
-
Research for Object-Oriented TM Image Segmentation and Information Extraction Technology
- Author(s):
-
GUO Peng1, XU Li-ping1, WANG Lin1, LIU Lin1, ZHANG Zheng-yong1, CHANG Cun2
-
1. College of Science, Shihezi University, Shihezi, Xinjiang 832003, China;
2. Xinjiang Institute of Ecology and Geography, Chinese Academy of Science Urumqi 830043, China
-
- Keywords:
-
object-orientation; Manas River Basin; classification technique
- CLC:
-
TP751
- DOI:
-
-
- Abstract:
-
Based on the theory of object-oriented remote sensing classification and ENV15.0 eCongition8.7 and ArcGIS10.0, we discussed key issues on multiscale segmentation and feature extraction, and extracted the feature information of the Manas River Basin. Results show that the best multiscale segmentation band weight were 0, 1, 2, 2, 1, 0, 1, both shape parameter and firmness was 0.7, segmentation scales were 60, 30 and 10. Feature types can be divided into woodland, grassland, water body, farmland, artificial surface, snow-coved land, saline-alkali soil and the other. The total accuracy of the classification results was 0.92 and Kappa was 0.91. It has been shown that the object-oriented classification method can effectively identify and extract the feature information of scale resolution images in the arid region.