image regression

Presently, unplanned changes of land use have become a major problem. Most land use changes occur without a clear and logical planning with little attention to their environment impacts.Since that landuse change occurring over large areas, remote sensing technology is an essential and useful tool for landuse change detection. In this study,after applying geometric and radiometric correction on landsat images of TM(1988) and ETM+(2001), five techniques of change detection have been used in 80470 hectare in the region of Daresher,Ilam province. These change detection techniques included Image regression, NDVI differencing, Principal component analysis (PCA(, Tasselled cap (KT) and post-classification comparison. In all these techniques, following standarizing maps,change direction has been determined.The accuracy of the results obtained by each technique was evaluated by comparison with post-classification method through Kappa coefficient calculation. According to the results, NDVI differencing and PC2 differencing showed the largest accuracy with Kappa coefficients of 0.667 and 0.655, respectively.However, Different change detection algorithms have their own merits and no single approach is optimal and applicable to all cases. In practice, several change detection techniques should be used to implement change detection, whose results are then compared to identify the best approach through visual or quantitative assessment.