Linear Impervious Feature Extraction by Entity based Image Analysis Approach using OLI/TIRS Imagery

This paper estimates the Linear Impervious feature of a transect, with Entity Based Image Analysis (EBIA) approach. The proposed approach incorporates the spatial entity as a major decision making unit to improve the linear feature extraction accuracy. Entity based analysis has better impervious mapping and further Entity based image analysis with indices improves the accuracy of impervious linear features. This approach is complement to the per pixel and other conventional methods, in which the decision is based on pixel alone and inherently incorporates the speckle noise.