Remote Sensing Image Classification ����������������������������������������������������������������������������
The classication of remotely sensed data has long attracted the attention of the remote sensing community because classication results are fundamental sources for many environmental and socioeconomic applications. Scientists and practitioners have made great efforts in developing advanced classication approaches and techniques for improving classication accuracy (Gong and Howarth 1992; Kontoes et al. 1993; Foody 1996; San Miguel-Ayanz and Biging 1997; Aplin, Atkinson, and Curran 1999; Stuckens, Coppin, and Bauer 2000; Franklin et al. 2002; Pal and Mather 2003; Gallego 2004; Lu and Weng 2007; Blaschke 2010; Ghimire, Rogan, and Miller 2010). However, classifying remotely sensed data into a thematic map remains a challenge because many factors, such as the complexity of the landscape under investigation, the availability of reference data, the selected remotely sensed data, image-processing and image classication approaches, and the analyst’s experiences, may affect classication accuracy. Many uncertainties or errors may be introduced into the classication results; thus, much effort should be devoted to the identication of these major factors in the image classication process and then to improving them. This chapter provides a brief overview of the major steps involved in the process of image classication, discusses the potential techniques for improving land-cover classication performance, and provides a case study for land use/cover classication in a moist tropical region of the Brazilian Amazon with Landsat thematic mapper (TM) imagery.