Land cover land use classification Using data mining techniques

Abstract

Natural catastrophes are a huge danger to life and belongings, and the level of damage has a direct relation with the susceptibility of the affected region. The number of times these natural catastrophes has occurred in the past few years has seen a manifold increase. The important catastrophes constitute tsunamis, floods, volcanic outbreaks, earthquakes etc. Floods fall under one of the most occurring and damaging natural disasters happening in the whole world. Abundant rainfall in a brief duration of time may lead to floods and the resultant discharge in the high river results in a huge amount of devastating effects. newlineClassically, field surveys or ground examinations were utilized for damage evaluation. The drawbacks of these methods are that when the catastrophe strikes, it is hard to get the crisis location. Remote sensing is a robust means of covering broad regions cost effectively and also it helps in detailed analysis of the situation during the disastrous hydrological event. newlineSatellite images are extensively utilized for monitoring urban improvement and land use cover variations at a medium or huge scale, to be useful in the better observation and interpretation of the development of urbanization and progress of the constant process of development. newlineThe highly popular techniques for flood detection include Support Vector Machines (SVM), Artificial Neural Networks (ANN), Convolutional Neural Networks (CNN) and decision tree classifier. These classification techniques are highly accurate for localized flooding circumstances, but, generalization capability is less with computation complexity being high, therefore, are not desirable for employing applications with multiple images. newline