Automatic detection of lineaments from LANDSAT images in United Arab Emirate (UAE)

摘要

Lineaments play tremendous role for preliminary exploration for characterization of geologic structure, oil exploration and ground water storages. Lineament identifications from remotely sensed data are required standard procedures. This work utilized a new geomatical tool for lineament identification which based on automatic detection algorithms. Integration between Laplacian, edge detection and Canny algorithms with different kernel window sizes is main component of new geomatical tool. The examination area is in mountainous areas of United Arab Emirate. The results show that: Canny algorithm with kernel window size of 3 x 3udpixels and line provides highly precisely lineaments information compared to Laplacian algorithm. Integration between Canny and edge detection algorithms illustrated highly accurate level within RMSE of ± 2 m.