پديد آورندگان :
محمدپور, محي الدين دانشگاه تهران - پرديس دانشكده هاي فني - دانشكده مهندسي معدن , بحرودي, عباس دانشگاه تهران - پرديس دانشكده هاي فني - دانشكده مهندسي معدن , ويسي, خالد اله دانشگاه صنعتي اصفهان - دانشكده مهندسي معدن
كليدواژه :
گسل مس پورفيري , استخراج اتوماتيك خطواره , تشخيص لبه و الگوريتم كني , تبديل هاف و جبال بارز
چكيده فارسي :
منطقه جبال بارز در كمربند ماگمايي اروميه – دختر قرار دارد و كانه زايي هاي مس پورفيري موجود در منطقه ارتباط زيادي با گسل‌ها و توده‌هاي نفوذي گرانيتوئيدي دارند. تعداد و محل شكستگي‌ها و خطواره‌ها، راهنماي مناسبي براي اكتشاف كانه‌زايي مس پورفيري مي‌باشند. در اين تحقيق از يك الگوريتم تمام اتوماتيك جهت استخراج خطواره‌ها از تصاوير ماهواره‌اي ASTER استفاده شد، بطوريكه مهمترين هدف اين الگوريتم، شناسايي دقيق لبه‌ها و نرخ خطاي كم است. براي بهبود تشخيص لبه، حالت بهينه الگوريتم كني و براي آشكارسازي خطواره‌ها، تبديل هاف اعمال شد، در نهايت پس از بررسي بر روي خطواره‌هاي تعيين شده، گسل‌هاي نهايي و مرتبط با كانه‌زايي شناسايي و نقشه آن تهيه شد. با توجه به نمودار رزدياگرام دو سيستم گسله بارزسازي شد كه سيستم گسله اصلي داراي امتداد شمال غربي - جنوب شرقي مي‌باشد و سيستم گسله با امتداد شمالي=جنوبي سهم ناچيزي از كل شكستگي‌هاي منطقه را تشكيل مي‌دهد. نقشه چگالي گسل‌ها، منطقه را از نظر كانه‌زايي مس پورفيري پر پتانسيل ارزيابي كرد. براي بررسي موفقيت‌سنجي تصاوير ASTER در استخراج عوارض خطي و گسل‌هاي منطقه، اين گسل‌ها با گسل‌هاي نقشه زمين‌شناسي مقايسه شد كه 63/05 درصد از آنها با گسل‌هاي زمين شناسي همپوشاني داشتند.
چكيده لاتين :
detection and extraction of geological lineaments are commonly applicable to the selection of right places for structures such as dams, bridges and roads (Charnpratheep et al., 1997; Mukherjee, 1999; Şener et al., 2011), water resources and hydrogeology studies (Chowdary et al., 2009; Krishnamurthy et al., 1996) and mineral exploration (Cooper, 2010; Lee et al., 2012). Due to the lack of access to some regions, as well as the need for time and cost to study large regions, researchers need to utilize more satellite imagery in geology and mineral exploration (Karimpour and Malekzadeh Shafaroudi, 2016). There are three methods, visual, semi-automatic, and automatic, for processing satellite images to detect lineaments (Rahnama and Gloaguen, 2014). Lineaments entirely determined in automatic methods. This method uses linear object algorithms such as Hough and Radon Transform algorithms. The success of automatic extraction of lineaments depends on the reliability and accuracy of the edge detection mechanism (Karantzalos and Argialas, 2006; Wladis, 1999). The automatic methods save time and cause the objectivity in improving the process of lineament extraction (Masoud and Koike, 2011; Rahnama and Gloaguen, 2014). Zhang et al. (2006) utilized the Radon transform for automatic extraction of lineaments (Zhang et al., 2006). Biswas and Sil (2012) developed the improved Canny edge detection algorithm to detect lines and development photos on the image (Biswas and Sil, 2012). Rahnama and Gloaguen (2014) analyzed lineaments on satellite images and implemented a Digital Elevation Model (DEM) using the Hough Transform algorithm (Rahnama and Gloaguen, 2014). Sanjay and Naoghare (2015) utilized the Canny method using two thresholds in the detection of lines (Sanjay and Naoghare, 2015). Among various methods to extract lineaments, those methods which are based on a combination of edge detector filters and Hough transform, yield the reliable and better results (Argialas et al., 2004; Fitton and Cox, 1998). The present research sought to use the optimized Canny algorithm and linear Hough transform as fully automatic methods to detect geological lineaments. This method applied on images of the ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) sensor on Jebal-e-Barez geological map. This Map is vital in terms of porphyry copper mineralization that is controlled mainly by tectonic fractures.
2-Methodology
The present study aimed to provide an optimized method for precise extraction of lineaments using satellite images. The proposed algorithm consisted of several implemented steps in MATLAB, as shown in Figure 1. First, the radiometric and geometric pre-processing carried out on satellite images, and then images were developed (become apparent) using the contrast stretching. The operation softened and reduced noise on images, and appropriate images obtained for Canny filtering. Edges of images were then detected, and the entry of false edges to the next stage was prevented by applying thresholds by Hysteresis method using Canny algorithms. Hough Transform was used to detect lines. Then tectonic lineaments developed at the post-processing stage; and the subsequent tectonic relationship with porphyry copper mineralization was then examined
