This assignment will include three
main types of classification, which compound: Two – band Feature Space
Classification, Unsupervised Classification, Supervised Classification. Land
Cover classification is one of the most important and typical applications of
remote sensing data. Land cover corresponds to the physical condition of the
ground surface. Remote sensing data can provide land cover information rather
than land use information, Initially the land cover classification system should
be established, which is usually defined as levels and classes.
Unsupervised
classification is one of two methods used to transform multispectral image data
into thematic information classes (supervised classification being the other).
This procedure typically assumes that imagery of a specific geographic are
gathered in multiple regions of the electromagnetic spectrum, for example
Landsat TM or SPOT XS multispectral satellite imagery. (Classification can also
be effective for other types of imagery. Please refer to an appropriate
reference text for complete information on classification.). In unsupervised
classification, the classification program automatically searches for natural
groupings or clusters of the spectral properties of pixels, and assigns each
pixel to a class based on initial clustering parameters you define.
Supervised
classification is one of two methods used to transform multispectral image data
into thematic information classes (unsupervised classification being the other).
This procedure typically assumes that imagery of a specific geographic are
gathered in multiple regions of the electromagnetic spectrum, for example
Landsat TM or SPOT XS multispectral satellite imagery. (Classification can also
be effective for other types of imagery. Please refer to an appropriate
reference text for complete information on classification.)
Objectives
Two
– band feature Space Classification use histogram examination, density
slicing and scatter diagram exploration techniques to perform simple
classifications, and determine the areas on the scatter plots that represent
distinctive and signification land cover types
Unsupervised
Classification purpose to analysis an unsupervised classification on the
image Aggregate classes as necessary to reduce confusion.
Supervised
Classification to performs supervised classification using the
parallelepiped minimum distance to means, and maximum likelihood techniques.
Develop training areas and create signature statistics for the training
area. Perform an accuracy assessment and examine the error matrix
Plotting
create a plot of the final supervised classification. Include as a minimum:
appropriate legend, projection information, a title, the author, date,
location, scale bar, and classification method.
Study area
In this case our lecturer provide east of java for
studying areas that we should received the data name e_java.ers file, then we
copy this file to our folder, and examine to perform procedure so on.
