Developing adaptive algorithm for automatic detection of geological linear features using RADARSAT-1 SAR data

Abstract

Synthetic aperture radar (SAR) has been recognized as a powerful tool for geological features detection. This work introduces a new approach for modification of adaptive Lee algorithm to semi-automatically detected geological features from RADARSAT-1 SAR data. This is based on modification of Lee algorithm formula by using 2D Gaussian convolution formula. The result shows that the new formula of Lee algorithm can be delineated lineament features in RADARSAT- 1 SAR data. The new formula based on 2D Gaussian performed better in semi-automatically detected lineament features as compared with Lee algorithm. The modification of Lee algorithm has a small standard deviation of 3.54 as compared to Lee algorithm.