Geographic Feature Extraction and Analysis using Geoinformatic Techniques

Abstract

newlineLand cover information is essential for a range of problems and themes in newlineearth observation sciences, such as environmental change, vegetation change, newlinehuman environment, especially urban climatology. These land cover newlineinformation is extracted from the multi-spectral and multi-temporal satellite newlineimagery dataset using the geoinformatic techniques. The remote sensing newlinesatellite imagery is more efficient than traditional geographic surveys and newlinestatistical data for extracting land cover features locally and globally. It is newlinetime saving and cost effective for mapping and analysis geographical changes newlinethan the traditional methods. The geoinformatics techniques is playing an newlineimportant role in the observation and analysis of all ecological and newlineenvironmental problems. The objective of the research study is to extract the newlinegeographical features of the land cover of the study area using geoinformatic newlinetechniques. The multi-spectral and multi-temporal Landsat TM/ETM+/OLI newlineimageries have been used in this research work as the primary dataset for newlinegeographic feature extraction. The image preprocessing methods are used to newlineenhance the information of multi-spectral imagery datasets, to classify more newlineaccurately land cover features. A new vegetation index method is proposed newlineto monitoring the health and growth of vegetation, orchards, and crops. It newlineclassifies vegetation and orchards more accurately than the other vegetation newlineindices using multispectral Landsat imagery datasets. The land cover feature newlineof the study area has a variety of geographical features which are divided into newlineseven classes such as orchards, vegetation, rangeland, agricultural land, newlineurban land, water bodies and watersheds. The supervised classifier MLC, newlineSVM, MD, and ANN have classified these land cover features into seven newlineland use and land cover (LU/LC) classes and produced the classified LU/LC newlinemaps. The confusion matrix method has used to accuracy assessment of newlineclassified LU/LC results. The post-classification change detection method newlinehas used on the resultant classified imagery data to detect the changes in land newlinecover features. The LU/LC map shows the change in land cover of the study newlinearea during the periods 1996 to 2107 due to the human activity and unplanned newlinedevelopment. The impact of unplanned urbanization and industrialization in newlinethe study area leads to problems of vegetation change, orchards degradation, newlineand water-body change. The LU/LC changes information is crucial to the newlineurban planner for monitoring, planning and decision-making and will be newlineuseful for planning future LU / LC