The purposes of this study are to extract lineaments from satellite images in order to contribute to the understanding of the faults. Landsat image is used for the analysis which is processed for both automated and manual extraction. During manual extraction four methods (filtering, PCA, band rationing and color composites) are used. Comparison of the two output maps indicated that manual extraction produced better results. Manually extracted lineament map is tested with the fault map of the area compiled from eight studies. The accuracy of the lineament map for the whole area is 38.69 % which increases to 50.28 % in the vicinity of North Anatolian Fault Zone (NAFZ). Evaluation of the length, density and orientation of the lineaments indicated that: a) there are fault zones in the area other than the NAFZ, b) Several fault segments are identified in the region which are absent in the fault map due the difficulty in mapping during the field studies / c) the dominant lineament trend is NE-SW (parallel to the NAFZ), however, a second trend is obvious in NW-SE direction.

The objective of the current research was to develop an automated algorithm with no or little user assistance for extraction of Subdural Grid Electrodes (SGE) from post-implant MRI scans for epilepsy surgery. The algorithm utilizes the knowledge about the artifacts created by Subdural Electrodes (SE) in MRI scans. Also the algorithm does not only extract individual electrodes, but it also extracts them as a SGE structures. Information about the number and type of implanted electrodes is recorded during the surgery [1]. This information is used to reduce the search space and produce better results. Currently, the extraction of SGE from post-implant MRI scans is performed manually by a technologist [1, 2, 3]. It is a time-consuming process, requiring on average a few hours, depending on the number of implanted SE. In addition, the process does not conserve the geometry of the structures, since electrodes are identified individually. Usually SGE extraction is complicated by nearby artifacts, making manual extraction a non-trivial task that requires a good visualization of 3D space and orientation of SGE in it. Currently, most of the technologists use 2D slice viewers for extraction of SGE from 3D MRI scans. There is no commercial software to perform the automated extraction task. The only algorithm suggested in the literature is [4]. The goal of the proposed algorithm is to improve the performance of the algorithm in [4]. As a goal, the proposed algorithm performs extraction of SGE not only for individual electrodes, but by applying geometric constraints on SGE.