No abstract

No abstract

No abstract

While classifying "built-up" pixels from satellite imagery, both machine learning & index based algorithms often misclassify "river sand" pixels as "built-up" ones due to the similarity in their spectral profiles. With the help of the spectral reflectance information in BLUE & GREEN bands of Landsat satellite imagery, this study has introduced a new index BRSSI (Built-Up & River Sand Separation Index) that efficiently reduce the misclassification between these two classes. The classification performance of the proposed index along with the same of Support Vector Machine (SVM) classifier have been reported for 3 study sites from different geographic locations across India. The results shows that average overall accuracy, F1 score and kappa (κ) coefficient for the developed index corresponding to selected study regions are 0.9763, 0.9767 & 0.9527 respectively. Though it has been noticed that SVM performs marginally better than BRSSI, it requires tuning of the parameters for optimum classification performance compared to BRSSI which is not only easy to implement but also computationally inexpensive. Also, visual representation of the classified images for entire study sites using additional filter of BRSSI ensures significant reduction of misclassification of "river sand" pixels as "built-up" class.

No abstract