Chin J Plant Ecol ›› 2005, Vol. 29 ›› Issue (6): 927-933.DOI: 10.17521/cjpe.2005.0121

• Research Articles • Previous Articles     Next Articles

LAND COVER CLASSIFICATION OF REMOTELY SENSED IMAGERY USING A METHOD BASED ON TOPOGRAPHICAL RESTRICTIVE FEATURES: A CASE STUDY OF THE JINHE WATERSHED

HONG Jun1,2, GE Jian-Ping2,*(), CAI Ti-Jiu2,3, NIE Yi-Huang4   

  1. 1 China Animal Husbandry & Veterinary Headstation, Beijing 100026, China
    2 Ministry of Education Key Laboratory for Biodiversity Science and Ecological Engineering & Institute of Ecology, Beijing Normal University, Beijing 100875, China
    3 College of Forestry, North-East Forest University, Harbin 150040, China
    4 Chinese Research Academy of Environment Science, Beijing 100012, China
  • Received:2004-10-13 Accepted:2005-03-29 Online:2005-10-13 Published:2005-09-30
  • Contact: GE Jian-Ping

Abstract:

Based on non-supervised classification methods, multi-temporal images, such as NOAA-AVHRR, SPOT-VEGETATION and MODIS series data, have been used to map regional, continental or global land cover patterns. Because of limitations of the classification method and spatial resolution of images, traditional non-supervised classification often results in many errors in the transition zones in which the spatial distribution of vegetation show segmental patterns. This problem can decrease the integral accuracy of the classification to a certain extent.
In this paper, a new method, the topographical restrictive features, is presented to classify remotely sensed images using the Jinhe watershed in the Loess plateau as a case study. First, 25 relatively independent classes were obtained using the non-supervised classification method with multi-temporal NDVI data derived from the first two bands of MODIS data. Every cell with 500 m spatial resolution of the classification determined by the non-supervised method was divided into 25 homogenous cells with 100 m spatial resolution. Then, topographical features were defined. Information about aspect, slope, elevation, river net structure, and vegetation regional characteristics, derived from 1:250 000 geographical spatial data, ETM+ image and yield data, were used to construct topographical restrictive features. Finally, the classification of every sub-cell was tested using the restrictive features and some cells were reclassified while maintaining their original classification. After the secondary classification, all cells were labeled by land cover type according to the land cover classification system, which was defined previously on the base of the IGBP land cover classification system and UMD system.
Using this new method, the accuracy of the land cover classification increased from 71.88%, in traditional non-supervised classification method, to 84.09%. In particular, the accuracy of cropland and urban type classification improved. The method of cell sub-division avoids the shortcomings of traditional classification methods owing to the coarse resolution of the image processing, and makes it more highly probable that land cover types are homogenous within cells. The introduction of topographical restrictive features decreases uncertainty of traditional fuzzy classification and provides more precise distinctive features to classify the fuzzy zone, and thus improves the accuracy of classification.

Key words: Image segmentation, Topographical restrictive features, Remotely sensed land cover classification, Jinhe watershed