WANG Xili, FU Li, MA Lei. Semi-supervised support vector regression model for remote sensing water quality retrieving[J]. Chinese Geographical Science, 2011, 21(1): 57-64.
Citation: WANG Xili, FU Li, MA Lei. Semi-supervised support vector regression model for remote sensing water quality retrieving[J]. Chinese Geographical Science, 2011, 21(1): 57-64.

Semi-supervised support vector regression model for remote sensing water quality retrieving

Funds:  Special Project of CAS-Russia, Ukraine and Belarus (2010)
More Information
  • Corresponding author: WANG Xi-Li
  • Received Date: 2010-05-27
  • Rev Recd Date: 2010-10-26
  • Publish Date: 2011-01-15
  • This paper proposes a semi-supervised regression model with co-training algorithm based on support vector machine, which retrieve water quality variables from SPOT5 remote sensing data. The model consists of two support vector regressors (SVR). Nonlinear relationship between water quality variables and SPOT5 spectrum are described by the two support vector regressors, and semi-supervised co-training algorithm for the SVRs is established. The model is used for retrieving concentrations of four representative water quality organic pollution indicators – permanganate index (CODmn), ammonia nitrogen (NH3-N), chemical oxygen demand (COD) and dissolved oxygen (DO) of the Weihe River in Shaanxi Province, China. The spatial distribution mappings for these variables over a part of the Weihe River in Shaanxi are also produced. SVR can implement any nonlinear mapping readily and semi-supervised learning can make use of both labeled and unlabeled samples. By integrating two support vector regressors and using semi-supervised learning, we provide an operational method when paired samples are limited. The results show that it is much better than the multiple statistical regression method, and can provide the whole water pollution conditions for management fast and can be extended to hyperspectral remote sensing applications.
  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Article Metrics

Article views(1407) PDF downloads(4) Cited by()

Proportional views
Related

Semi-supervised support vector regression model for remote sensing water quality retrieving

Funds:  Special Project of CAS-Russia, Ukraine and Belarus (2010)
    Corresponding author: WANG Xi-Li

Abstract: This paper proposes a semi-supervised regression model with co-training algorithm based on support vector machine, which retrieve water quality variables from SPOT5 remote sensing data. The model consists of two support vector regressors (SVR). Nonlinear relationship between water quality variables and SPOT5 spectrum are described by the two support vector regressors, and semi-supervised co-training algorithm for the SVRs is established. The model is used for retrieving concentrations of four representative water quality organic pollution indicators – permanganate index (CODmn), ammonia nitrogen (NH3-N), chemical oxygen demand (COD) and dissolved oxygen (DO) of the Weihe River in Shaanxi Province, China. The spatial distribution mappings for these variables over a part of the Weihe River in Shaanxi are also produced. SVR can implement any nonlinear mapping readily and semi-supervised learning can make use of both labeled and unlabeled samples. By integrating two support vector regressors and using semi-supervised learning, we provide an operational method when paired samples are limited. The results show that it is much better than the multiple statistical regression method, and can provide the whole water pollution conditions for management fast and can be extended to hyperspectral remote sensing applications.

WANG Xili, FU Li, MA Lei. Semi-supervised support vector regression model for remote sensing water quality retrieving[J]. Chinese Geographical Science, 2011, 21(1): 57-64.
Citation: WANG Xili, FU Li, MA Lei. Semi-supervised support vector regression model for remote sensing water quality retrieving[J]. Chinese Geographical Science, 2011, 21(1): 57-64.

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return