Abstract
Lineaments observed from a sattelite imagery are characterized by their positions, the directions and the lengths on a two dimensional space. By analysing their dominant direction, the frequency and the total length on a certain direction, systematic and statistical characteristics of groups of lineaments can be evaluated.
In this study, Hough transform was proposed as an efficient method for lineament analyses. The performance of this method was investigated using a lineament distribution map extracted from a Landsat imagery. As a result, Hough transform was proved to be very useful tool to make a Rose Diagram of trend-total length with a high degree of accuracy. The correlation coefficient between Rose Diagrams by the Hough transform and those by the actual data is more than 0.98.
Fourier transform is a similar method for lineament analyses. The performance difference between these two transform methods was also examined and compared.