2.3 Image Classification 2.3.1 Supervised classification To produce the Land cover maps, supervised classification algorithm was used.

Based on the ancillary data and information gathered from the field, the land cover types in the study area were generated using the LCCS software package (FAO, 2005). In this research, the FAO Land Cover Classification System was used. 2.2.3 Classification scheme Definition of a classification scheme is an initial step in any classification project. This information was used in image classification and accuracy assessment. Also, to obtain historical land cover data, the local people were interviewed at different locations to determine whether or not a particular land cover category existed at the different time periods under consideration. Field verification was then carried out to determine coordinates of the sites with GPS as well as the nature of land cover at such sites. This exercise helped in the pre-determination of training/test sites. This allowed natural spectral clusters to be defined with high level of objectivity and reliability. 2.2.2 Field data Prior to image classification, cluster analysis was carried out using satellite imageries of the different time periods. These imageries were photographs of the condition of the environment before and after Akwa Ibom State was created in 1987.

Consequently, the orthorectified Landsat TM satellite data for 1986 and NIGERSAT-1 imagery for 2007 were acquired and used for land cover mapping and change detection/analysis. The 28.5m resolution landsat TM satellite imagery has become the standard. The coverage area of the satellite imageries that were used is shown in yellow color on the map 2.2 Data acquisition 2.2.1 Remote sensing Data Land cover change analysis and mapping requires high spatial and temporal resolution satellite data. Location of Akwa Ibom State, Nigeria One of the constraints to this study was the lack of current cloud free satellite imagery covering the entire State.

Descargar SWF Jogja Map