Work

Sub-pixel Image Classification of Hyper-spectral Data for Vegetation and Soil Mappingin Semi-arid Environment

Year: 2009
Type: article
Abstract: The HyMap hyper-spectral data was used to classify photosynthetic\nvegetation (PV), non-photosynthetic vegetation (NPV), and exposed soils in a semiarid\nsavannah environment of McKinlay, northern Queensland, and Australia. This\nstudy aimed to understandhow effective the sub-pixel classificationapproach applied\non hyper-spectral data to distinguish the vegetation and soil features in semi-arid\nenvironment. In contrast to the per-pixel approach this approach treats the pixel\nvalue as reflectance sum of its composite features, and shows its component\nabundance. The most commonly used sub-pixel classification technique was used in\nthis research, namely Linear Spectral Unmixing (LSU). End members were used as\nthe input class, and the result was compared with the standard maximum likelihood\nclassification (MLC) using post-classification comparison method The result of this\nstudy shows that LSU produced a patchy distribution of classes throughout the\nimage. The brown soil tends to be over-estimated with respect to other classes. PV\nfeatures were relatively well-mapped compare to other classes. NPV features have\nproblem with domination of exposed soil reflectance. This is equivalent to the\nprevious studies result that background soil dominates the spectral reflectance in\nthis environment. According to the qualitative accuracy assessment, LSU has\nhigher accuracy in representing PV and NPV compare to the traditional MLC\nclassification.
Language: English
Open Access status: gold
OpenAlex ID: W1937677300
FWCI:
Cites: 8
Cited by:
Related to: 10