No abstract

ABSTRACT:Shorelines are complex ecosystems and highly important socio-economic environments. They may change rapidly due to both natural and human-induced effects. Determination of movements along the shoreline and monitoring of the changes are essential for coastline management, modeling of sediment transportation and decision support systems. Remote sensing provides an opportunity to obtain rapid, up-to-date and reliable information for monitoring of shoreline. In this study, approximately 120 km of AntalyaKemer shoreline which is under the threat of erosion, deposition, increasing of inhabitants and urbanization and touristic hotels, has been selected as the study area. In the study, RASAT pansharpened and SENTINEL-1A SAR images have been used to implement proposed shoreline extraction methods. The main motivation of this study is to combine the land/water body segmentation results of both RASAT MS and SENTINEL-1A SAR images to improve the quality of the results. The initial land/water body segmentation has been obtained using RASAT image by means of Random Forest classification method. This result has been used as training data set to define fuzzy parameters for shoreline extraction from SENTINEL-1A SAR image. Obtained results have been compared with the manually digitized shoreline. The accuracy assessment has been performed by calculating perpendicular distances between reference data and extracted shoreline by proposed method. As a result, the mean difference has been calculated around 1 pixel.

ABSTRACT:Coastlines are important features for water resources, sea products, energy resources etc. Coastlines are changed dynamically, thus automated methods are necessary for analysing and detecting the changes along the coastlines. In this study, Sentinel-1 C band SAR image has been used to extract the coastline with fuzzy logic approach. The used SAR image has VH polarisation and 10x10m. spatial resolution, covers 57 sqkm area from the south-east of Puerto-Rico. Additionally, radiometric calibration is applied to reduce atmospheric and orbit error, and speckle filter is used to reduce the noise. Then the image is terrain-corrected using SRTM digital surface model. Classification of SAR image is a challenging task since SAR and optical sensors have very different properties. Even between different bands of the SAR sensors, the images look very different. So, the classification of SAR image is difficult with the traditional unsupervised methods. In this study, a fuzzy approach has been applied to distinguish the coastal pixels than the land surface pixels. The standard deviation and the mean, median values are calculated to use as parameters in fuzzy approach. The Mean-standard-deviation (MS) Large membership function is used because the large amounts of land and ocean pixels dominate the SAR image with large mean and standard deviation values. The pixel values are multiplied with 1000 to easify the calculations. The mean is calculated as 23 and the standard deviation is calculated as 12 for the whole image. The multiplier parameters are selected as a: 0.58, b: 0.05 to maximize the land surface membership. The result is evaluated using airborne LIDAR data, only for the areas where LIDAR dataset is available and secondly manually digitized coastline. The laser points which are below 0,5 m are classified as the ocean points. The 3D alpha-shapes algorithm is used to detect the coastline points from LIDAR data. Minimum distances are calculated between the LIDAR points of coastline with the extracted coastline. The statistics of the distances are calculated as following; the mean is 5.82m, standard deviation is 5.83m and the median value is 4.08 m. Secondly, the extracted coastline is also evaluated with manually created lines on SAR image. Both lines are converted to dense points with 1 m interval. Then the closest distances are calculated between the points from extracted coastline and manually created coastline. The mean is 5.23m, standard deviation is 4.52m. and the median value is 4.13m for the calculated distances. The evaluation values are within the accuracy of used SAR data for both quality assessment approaches.

Çevrenin korunması ve sürdürülebilir kıyı geliştirme hedeflerine ulaşmak için kıyı alanlarının izlenmesi gerekmektedir. Doğal çevre yönetimi, afet yönetimi, kıyı erozyonu incelemeleri, katı madde taşınımı ve kıyı morfodinamiklerinin modellemesi gibi farklı alanlarda kıyı çizgisi yaygın olarak kullanıldığından, kıyı çizgilerinin özellikle uydu görüntülerinden çıkarılması için çeşitli teknikler geliştirilmiştir. Rastgele Orman algoritması geliştirilen bu tekniklerden bir tanesidir. Rastgele Orman Algoritması, karar ağaçlarına dayanan bir makine öğrenme metodudur. Karar ağaçları, eğitim verilerinin sınıflarını analiz eder ve test verilerinin hangi sınıfa ait olduğunu eğitim verilerinden çıkarttığı kurallara göre belirler. Bu çalışmada, Terkos Gölü ve gölün hemen yakınında yer aldığı bölgedeki Karadeniz sahilindeki kıyı çizgileri, 22 Temmuz 2016 tarihinde alınan Landsat-8 uydu görüntüsüne Rastgele Orman sınıflandırma yöntemi kullanılarak çıkarılmıştır. Uygulamada öncelikle uydu görüntüsüne ön işleme uygulanmıştır. Rastgele Orman algoritması ile sınıflandırma işlemi için MATLAB platformu kullanılmıştır. Rastgele Orman algoritması çalışma bölgelerine ait görüntülerin farklı bant setlerine uygulanarak, sınıflandırma işlemi gerçekleştirilmiştir. Sınıflandırma sonucunda kara ve su sınıfları olmak üzere ikili görüntüler elde edilmiştir. Uygulanan bant setleri NIR, R-G-B ve R-G-B-NIR'dir. Terkos Gölü'ne ait elde edilen kıyı çizgilerinin doğruluklarını analiz etmek için elle sayısallaştırılan kıyı çizgileri referans alınarak alansal ve kıyı çizgisi bazında karşılaştırılmıştır. Sonuçlara bakıldığında, kıyı çizgisi çıkarımında yakın kızıl ötesi bandını içeren görüntülerin en az hatalı sonucu verdiği görülmüştür.

ABSTRACT:Coastlines are important features for water resources, sea products, energy resources etc. Coastlines are changed dynamically, thus automated methods are necessary for analysing and detecting the changes along the coastlines. In this study, Sentinel-1 C band SAR image has been used to extract the coastline with fuzzy logic approach. The used SAR image has VH polarisation and 10x10m. spatial resolution, covers 57 sqkm area from the south-east of Puerto-Rico. Additionally, radiometric calibration is applied to reduce atmospheric and orbit error, and speckle filter is used to reduce the noise. Then the image is terrain-corrected using SRTM digital surface model. Classification of SAR image is a challenging task since SAR and optical sensors have very different properties. Even between different bands of the SAR sensors, the images look very different. So, the classification of SAR image is difficult with the traditional unsupervised methods. In this study, a fuzzy approach has been applied to distinguish the coastal pixels than the land surface pixels. The standard deviation and the mean, median values are calculated to use as parameters in fuzzy approach. The Mean-standard-deviation (MS) Large membership function is used because the large amounts of land and ocean pixels dominate the SAR image with large mean and standard deviation values. The pixel values are multiplied with 1000 to easify the calculations. The mean is calculated as 23 and the standard deviation is calculated as 12 for the whole image. The multiplier parameters are selected as a: 0.58, b: 0.05 to maximize the land surface membership. The result is evaluated using airborne LIDAR data, only for the areas where LIDAR dataset is available and secondly manually digitized coastline. The laser points which are below 0,5 m are classified as the ocean points. The 3D alpha-shapes algorithm is used to detect the coastline points from LIDAR data. Minimum distances are calculated between the LIDAR points of coastline with the extracted coastline. The statistics of the distances are calculated as following; the mean is 5.82m, standard deviation is 5.83m and the median value is 4.08 m. Secondly, the extracted coastline is also evaluated with manually created lines on SAR image. Both lines are converted to dense points with 1 m interval. Then the closest distances are calculated between the points from extracted coastline and manually created coastline. The mean is 5.23m, standard deviation is 4.52m. and the median value is 4.13m for the calculated distances. The evaluation values are within the accuracy of used SAR data for both quality assessment approaches.