1M.Saranya, 2S. Kalaiselvi, 3Dr.K.G. Srinivasagan
1 PG-Scholar, Department of CSE-PG, National Engineering College Kovilpatti,Tamilnadu,India.
2 Assistant Professor, Department of CSE-PG National Engineering College Kovilpatti,Tamilnadu,India.
3 Professor and Head , Dept of CSE-PG, National Engineering college,Kovilpatti,Tamilnadu,India.
1 PG-Scholar, Department of CSE-PG, National Engineering College Kovilpatti,Tamilnadu,India.
2 Assistant Professor, Department of CSE-PG National Engineering College Kovilpatti,Tamilnadu,India.
3 Professor and Head , Dept of CSE-PG, National Engineering college,Kovilpatti,Tamilnadu,India.
Land cover of the earth’s
land surface has been changing since time immemorial and is likely to continue
to change in the future. These changes are occurring at a range of spatial
scales from local to global and at temporal frequencies of days to millennia.
Both natural and anthropogenic forces are responsible for the change. Natural
forces such as continental drift, glaciations, flooding, tsunamis and
anthropogenic forces such as conversion of forest to agriculture, urban sprawl,
and forest plantations have changed the dynamics of land-use/land-cover types
throughout the world. The main intention of the classification is to group the
pixels in the image into form the several land cover classes, or “themes”. And
then the categorized data is used to produce the thematic map of land cover classes.
Image classification is one of the most important parts in image processing.
Image segmentation is the process of dividing the images into regions with
similar attributes. The proposed research consists of three phases. First
phase, the image segmentation is done by Self Organizing Map (SOM). Second
phase, feature extraction is done by Gray Level Coocurrence Matrix methods.
Finally the classification is done by fuzzy rule based classification. The
final output is land use and land cover map, with labels of corresponding
classes.
