آشكار سازي تغييرات سطوح نفوذ ناپذير در محدوده كلان شهر اصفهان

چكيده انگليسي :

Change detection of impervious surfaces in Isfahan metropolitan area Azadeh Zaeri Amirani a zaeri@na iut ac ir 31 10 2010 Department of Natural Resources Isfahan University of Technology Isfahan 84156 83111 Iran Degree M Sc Language Farsi Ali Reza Soffianian Soffianian@cc iut ac ir Abstract Since urbanization is an inevitable process we can direct urban growth in the most suitable direction by land use planning so it can provide dwellers requirements and also maintain natural resources suburbans and agricultural areas Also the impervious surface has been used as an indicator of urbanization This surfaces effect on urban ecology such as urban heat island affect restrictions on water purification and infiltration in soil increased pollution runoff and increased erosion in the downstream watershed In recent years remote sensing data and geographic information system GIS have been widely applied in identifying and analyzing land use and land cover and impervious surface change Accurate change detection of earth s surfaces provides better understanding of relationships and interactions between human and natural phenomena for prefect resource management There are a lot of methods for land cover change detection such as image rationing image difference change vectore analysis image regression composite analysis and post classification Post classification is one of the most effective change detection methods The objective of this study is change detection of Isfahan land use land cover from 1972 to 2008 MSS images in 1972 and 1987 TM and ETM images of Landsat satellite in 1995 and 2001 and LISS III and AWiFS images of IDS 1D in 2008 were used in current study All of satellite images were rectified by using first degree polynomial equation and nearest neighbor sampling RMSe of TM ETM MSS in 1972 MSS in 1987 LISS III and AWiFS images were found 0 45 0 42 0 6 0 58 0 4 and 0 56 pixels respectively indicating that the spatial error is expected to be less than one pixel To make the best false color composite we used Optimal Index Factor OIF to identify best bands with minimum correlation and maximum variance To classify images Maximum Likelihood Fuzzy ARTMAP Multi Layer Perceptron neural network classifier Kohonen s Self Organizing Map and tree decision classification was used Finally land cover maps were provided with 5 classes that including water river residential area bare plant cover and roads Error matrix was calculated for accuracy assessment of maps Total accuracy of produced land cover maps from Layer Perceptron neural network classification method were maximum ranges Total accuracy from TM ETM MSS in 1972 MSS in 1987 LISS III and AWiFS was 93 23 93 78 91 15 90 79 93 29 and 92 86 respectively Then changes matrix was produced by comparing land cover map of each year with next year s map For change detection post classification comparison and radiometric method was performed Because in this study we used different imaging sensor and radiometric method for change detection it was necessary to remove influence of different solar altitudes different angles different meteorologic condition and different imaging sensor that leads to misclassify Therefore for pre classification techniques images were radiometrically normalized first In this study radiometric change detection including image differencing image rationing and image regression were used In addition land cover changes using change vector analysis performed Generally results show that radiometric methods for land cover change detection in urban area with high variety have lower accuracy than post classification but among this methods image differencing contains more accuracy than other Change detection results show that impervious surfaces area increased from 8434 ha to 1951 increase 15321 ha in 2008 Key words Impervious surfaces Image classification Multi Layer Perceptron Maximum Likelihood Change detection Post classification pre classification methods

زائري اميراني، آزاده

آشكار سازي تغييرات سطوح نفوذ ناپذير در محدوده كلان شهر اصفهان

طبقه بندي تصوير , پرسپترون چند لايه , حداكثر احتمال , مقايسه پس از طبقه بندي و روش هاي پيش از طبقه بندي