Papers by Umut Gunes Sefercik
Assessing the Physical and Chemical Characteristics of Marine Mucilage Utilizing In-Situ and Remote Sensing Data (Sentinel-1, -2, -3)
Turkish journal of electrical engineering and computer sciences/Elektrik, Feb 7, 2024
The utilization of remote sensing products for vehicle detection through deep learning has gained... more The utilization of remote sensing products for vehicle detection through deep learning has gained immense popularity, especially due to the advancement of unmanned aerial vehicles (UAVs). UAVs offer millimeter-level spatial resolution at low flight altitudes, which surpasses traditional airborne platforms. Detecting vehicles from very high-resolution UAV data is crucial in numerous applications, including parking lot and highway management, traffic monitoring, search and rescue missions, and military operations. Obtaining UAV data at desired periods allows the detection and tracking of target objects even several times during a day. Despite challenges such as diverse vehicle characteristics, traffic congestion, and hardware limitations, the detection task must be executed swiftly and accurately. This study successfully achieved automated detection and instance segmentation of parked and moving vehicles across a large university campus by employing the robust learning capabilities of the You Only Look Once version 7 (YOLOv7) deep learning algorithm. The generation of an ultrahigh-resolution orthomosaic of the university campus was accomplished through photogrammetric processing, employing 20-megapixel aerial images obtained from RGB UAV flights with polygonal nadir-view and bundle-grid oblique-view imaging geometries. The vehicle dataset was created by cropping image patches containing vehicle objects from the orthomosaic and manually labeling the boundaries of the vehicle targets using the LabelMe annotation tool. After expanding the dataset by applying data augmentation, the YOLOv7 algorithm was trained and tested using the transfer learning approach. The accuracy metric of precision, recall, and mAP@0.50 scores for the bounding boxes and masks of vehicles were estimated as 99.79, 97.54, and 99.46%, respectively. In addition, the robustness of the trained algorithm was also tested on a short video and (> 80%) prediction scores were achieved.
UZALCBS 2022 Sempozyumu
İnsansız hava aracı (İHA) sistemleri düşük uçuş irtifasının getirisi olan birçok avantaj sayesind... more İnsansız hava aracı (İHA) sistemleri düşük uçuş irtifasının getirisi olan birçok avantaj sayesinde son yıllarda bilimsel çalışmalarda sıklıkla kullanılır hale gelmiştir. İHA sistemlerinin popüler hale gelmesinde, uygulanabilir genişlikte çalışma alanlarında, yüksek çözünürlüklü verinin geleneksel hava kaynaklı uzaktan algılama yöntemlerine nazaran daha düşük maliyetle ve periyodik olarak elde edilebilmesi önemli bir faktördür. Optik İHA, yüksek çözünürlüklü dijital kameralar ile farklı irtifalarda bindirmeli çekim geometrilerinde çalışabilme imkânı sayesinde çok çeşitli yükseklik ve formlara sahip objelere ait hassas ve orjinal renkte üç boyutlu (3B) modellerin üretimine olanak vermesi nedeniyle en çok tercih edilen İHA teknolojisi konumundadır. Ancak pasif uzaktan algılama prensibi ile işletilmesi, güneş ışığına bağımlılığı ve gündüz saatleri dışında operasyonel olmama durumunu beraberinde getirmektedir. Yanı sıra, geniş dalga boylu radar ya da lazer sistemlerde mevcut bitki penetrasyon kabiliyetinin olmayışı özellikle orman dokusu altı bölgelerde 3B tasvirde ciddi problemler yaratmaktadır. Ek olarak, hava fotoğraflarının eşlenmesi ve yöneltmesinde, görüntü alım geometrisi ve çalışma sahasındaki düşük korelasyonlu arazi örtüsünün meydana getirdiği sınırlamalar üretilen yoğun nokta bulutları ve 3B modeller üzerinde hata ve eksik tasviri beraberinde getirmektedir. Bu çalışmada, 20 MP Sony Exmor RGB kamerasının avantajları nedeniyle kullanıcılar tarafından en sık tercih edilen optik ticari İHA'lardan biri haline gelmiş DJI Phantom 4 Pro V2.0 verilerinden 3B model üretiminde dikkat edilmesi gereken problematik noktalar ve çözüm önerileri sunulmuştur. Çalışmada, SfM (hareketten yapı oluşturma) tabanlı görüntü eşleştirme yazılımlarında 3B topoğrafik tasvir ve obje tasvirinde kullanılan parametrelerin neden olduğu farklı etkiler ortaya konmuştur.
Comprehensive comparison of VHR 3D spatial data acquired from IKONOS and TerraSAR-X imagery
Advances in Space Research, 2013
ABSTRACT With the launch of very high resolution (VHR) optical and synthetic aperture radar (SAR)... more ABSTRACT With the launch of very high resolution (VHR) optical and synthetic aperture radar (SAR) remote sensing satellites such as IKONOS and TerraSAR-X, a new era has begun in 3D spatial data acquirement. IKONOS provided the first VHR data is still being preferred for many remote sensing applications. TerraSAR-X is considered as a revolution in SAR imaging as a result of 1 m resolution imaging capability. The imaging principles of these satellites are quite different and include advantages and disadvantages that have considerable effects on the quality of acquired 3D spatial data.
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Sep 7, 2012
Photogrammetry and remote sensing have become indispensable in today's surveying with the emergin... more Photogrammetry and remote sensing have become indispensable in today's surveying with the emerging technologies. Some difficulties are being experienced in applications due to the incredible advances in imaging systems and increasing complexity of carrier platforms, used, in parallel with them. In this context, the most agreeable approach is the achievement of desired precision proportional with the requirements in shorter time span and the reduction of cost utilizing the most suitable mapping platform with basic systems. Towards these aims, a model aircraft, containing a non-metric camera, capturing images for mapping purposes, is considered to be employed in this study. Thus, this will lead to emerge a low cost imaging system. Within the study, besides using a high resolution digital non-metric camera and a carrier model aircraft, the previously produced maps of test field, derived from aerial photos, were used to conduct the analysis. The scope of this study is the integration of a model aircraft and a non-metric camera for mapping purposes to reduce very costly and time consuming photogrammetric map production for narrower areas. Also this will give rise to a method for the achievement of high cost aims with lower budgets. Using a model aircraft, instead of a real one, gives us an upper hand in repeating measurements anytime, desired, so that analyses can be made for the works that undergoes temporal changes. According to the aims of the study, a stereo model has been generated for the extraction of details from aerial photos, taken by the non-metric camera. Estimated results have been compared with 1/1000 scaled maps from photogrammetry and figured out their validity for using in narrower areas with high accuracy.
Roof Modelling Potential of Unmanned Air Vehicle Point Clouds with Respect to Terrestrial Laser Scanning
cosp, Jul 1, 2016
Marmara Denizindeki Müsilaj Olayının Uzaktan Algılama Teknolojileri ile Tespiti ve İzlenmesi
Turkish Academy of Science eBooks, 2021
Küresel ısınmanın karasal alanlarda olduğu gibi denizlerde de doğal dengeyi bozan etkileri bulunm... more Küresel ısınmanın karasal alanlarda olduğu gibi denizlerde de doğal dengeyi bozan etkileri bulunmaktadır. Adriyatik ve Tiren denizlerinde 200 yılı aşkın süredir sıklıkla yaşanan müsilaj olayları, Marmara Denizinde ilk kez 2007 yılında gözlemlenmiş, 2021 yılı mayıs ayından itibaren ise tekrar yoğun şekilde ortaya çıkarak doğal bir felakete dönüşmüştür. Deniz yüzeyinde geniş alanlar kaplayan müsilaj, deniz altındaki canlı hayatını da ciddi şekilde tehdit etmektedir. Müsilajın alansal büyüklüğünün, yoğunluğunun, zamansal hareketinin ve deniz suyu ile arasındaki ısı farkının takibi ve sahada yürütülen temizlik çalışmalarının etkilerinin tespiti için uzaktan algılama teknolojileri etkin bir çözümdür. Periyodik olarak üretilen müsilaj dağılım haritaları, deniz yüzeyindeki toplam müsilaj hacminin hesaplanması, rehabilitasyon çalışmalarının yürütülmesi ve alınacak önlemlerin belirlenmesi başta olmak üzere birçok faaliyet için temel altlık konumundadır. Bu çalışmada, Sentinel-2 uydu görüntüsünün dört spektral bandı arasındaki ilişkiyi kullanan ‘müsilaj indeksi’ ile müsilaj kaplı alanların tespiti hedeflenmiştir. Bu indeksten faydalanılarak 14, 19 ve 24 Mayıs 2021 tarihli uydu görüntüleri için müsilaj indeks haritaları oluşturulmuştur. Sentinel-2 için tanımlanan müsilaj indeksi ile müsilaj oluşumlarının hızlı bir şekilde tespit ve analizinde önemli bir veri seti oluşturulabileceği görülmüştür. Ayrıca 17 Mayıs 2021 tarihli Landsat-8 termal uydu görüntüsü yardımıyla temiz deniz yüzeyi ve müsilaj oluşumlarının yüzey sıcaklıkları analiz edilmiştir. Sonuç olarak, müsilaj alanlarının yüzey sıcaklığının ortalama deniz yüzey sıcaklığına göre 3°C’ye kadar daha yüksek olduğu tespit edilmiştir. Elde edilen bulgular, uzaktan algılamanın erken uyarı ve hazırlık çalışmalarındaki etkinliğini ve afet riski yönetiminde vazgeçilmez bir araç olduğunu ortaya koymuştur.
Optical Engineering, Feb 4, 2013
Advancements in the geometric resolution of space images have improved the conditions for generat... more Advancements in the geometric resolution of space images have improved the conditions for generations of large-scale topographic maps. Using WorldView-1, WorldView-2, and GeoEye-1, images can now be captured from space with a 0.5 m ground sampling distance (GSD). Geometric accuracy and information content are the most significant components of mapping from space images. Depending on the resolution, image quality, and shadows, the identification and classification of ground objects may prove challenging. In this research, the geometric accuracy and information content, of panchromatic WorldView-1 images, were analyzed by covering parts of Istanbul and Zonguldak in Turkey. Each of these locations has various topographic characteristics. For the orientation and investigation of the geometric accuracies of images, a number of ground control points (GCPs) were developed as independent checkpoints. Based on bias-corrected rational polynomial coefficients with one GCP, a standard deviation of independent checkpoints on the range of one GSD was obtained. The information content of images was analyzed by mapping all buildings, in both test areas, and comparing the results with reference 1∕5000 scaled topographic maps. The results verified that the WorldView-1 images can be utilized for generating and updating 1∕5000 scaled topographic maps of urban areas.
Piksel Tabanlı Evrişimsel Sinir Ağları Kullanılarak Müsilaj Oluşumlarının Tespiti ve İzlenmesi: İzmit Körfezi Örneği
Marmara Denizi’ndeki Müsilaj Oluşumlarının Çok Zamanlı Optik ve Termal UyduGörüntülerinden Makine Öğrenme Algoritması ile Tespiti ve Analizi
Türkiye fotogrametri dergisi, Dec 29, 2020
Lokal hareket işaretleri ile iki boyutlu görüntü dizilerinin birleşiminden üç boyutlu (3B) yapıla... more Lokal hareket işaretleri ile iki boyutlu görüntü dizilerinin birleşiminden üç boyutlu (3B) yapıları kestirebilmek için geliştirilmiş Hareketten Yapı (Structure From Motion, SFM) fotogrametrik görüntü eşleştirme algoritması, yeni nesil ve yaygın olarak kullanılan bulut tabanlı görüntü eşleştirme yazılımlarının temel prensibidir. Bu yazılımlar ortak prensipte çalışmasına rağmen, kullanıcı tarafından müdahele edilemeyen gömülü parametrelerine bağlı olarak 3B sonuç ürünleri farklı özellikler ve distorsiyonlar içermektedir. Bu çalışmada, Zonguldak Bülent Ecevit Üniversitesi Çaycuma Kampüsü'nde insansız hava aracı (İHA) ile elde edilen yüksek çözünürlüklü hava fotoğraflarından VisualSFM, Agisoft ve Pix4D SFM tabanlı yeni nesil görüntü eşleme yazılımları kullanılarak eş grid aralıklı 3B dijital yüzey modelleri (DYM) üretilmiştir. Üretilen DYM'ler kapsamlı bir şekilde değerlendirilmiş ve Agisoft DYM'si referans olarak kullanılarak DYM'ler görsel ve istatistiksel yaklaşımlarla karşılaştırılmıştır. Standart sapma ve normalize medyan mutlak sapma temelinde elde edilen sonuçlar, analiz edilen SFM tabanlı yazılımların artılarını ve eksilerini açıkça ortaya koymuştur.
Yoğun, yüksek konum doğruluklu, düşük maliyetli ve periyodik nokta bulutları elde etme ve doğruda... more Yoğun, yüksek konum doğruluklu, düşük maliyetli ve periyodik nokta bulutları elde etme ve doğrudan üç boyutlu dijital yüzey ve arazi modellerini sağlayabilme gibi avantajları sayesinde Hava Kaynaklı Lazer Tarama (ALS), haritacılık başta olmak üzere yeryüzü ve ona dair verileri inceleyen bilim alanlarında oldukça sık kullanılır hâle gelmiştir. ALS, özellikle mali açıdan güçlü Avrupa ülkeleri ve ABD'de, aktif algılamanın bulut ve orman penetrasyon yeteneği gibi avantajlarından faydalanarak pasif algılama prensipli yani güneşin aydınlatmasına bağımlı fotogrametriye alternatif bir yöntem olarak kısa sürede kabul görmüştür. Bununla birlikte, ALS teknolojisi, ihtiyaç duyduğu uçak, hava kaynaklı lazer tarayıcı, Global Navigasyon Uydu Sistemleri (GNSS), Atalet Ölçüm Birimi (IMU) ve yüksek kapasiteli iş istasyonları gibi ekipmanların yüksek maliyetinden ötürü az gelişmiş veya gelişmekte olan ülkelerde yeterince kullanılamamıştır. Finansal gücü yeterli olmayan az gelişmiş ve gelişmekte olan ülkelerde de ALS tekniğinin kullanılabilmesini sağlamak için araştırma ekibimiz, Ulusal bir Araştırma ve Geliştirme (Ar-Ge) Projesi kapsamında, düşük maliyetli ALS kabiliyetli gelişmiş bir insansız hava aracı (İHA) üretmiştir. Cihaz, yeni dizayn ALS gimbal ve gerçek zamanlı kinematik konumlandırma gibi gelişmiş özelliklerle yüksek doğruluk ve çözünürlükte ALS verileri sağlayabilmektedir. Bu çalışmada, üretilen ALS, İHA ve verilerinden elde edilen dijital yüzey modelinin (DYM) kalitesi, açık, orman ve bina arazi sınıflarını içeren Bülent Ecevit Üniversitesi Merkez Kampüsü'ndeki bir test alanında görsel ve istatistiksel nokta ve model bazlı yaklaşımlarla ortaya konmuştur. Nokta bazlı analizlerde, araziden toplanan yer kontrol noktaları kullanılırken model bazlı değerlendirmelerde yersel lazer taramadan elde edilen referans DYM'den faydalanılmıştır. Elde edilen sonuçlar, üretilen hava kaynaklı lazer tarayıcının veri performansının uluslararası ALS veri standartlarını fazlasıyla sağlayabildiğini ortaya koymuştur.
Journal of The Indian Society of Remote Sensing, Sep 1, 2009
By Interferometric Synthectic Aperture Radar (InSAR), during the Shuttle Radar Topography Mission... more By Interferometric Synthectic Aperture Radar (InSAR), during the Shuttle Radar Topography Mission (SRTM) height models have been generated, covering the earth surface from 56° south to 60.25°n orth. With the exception of small gaps in steep parts, dry sand deserts and water surfaces, the free available US C-band data cover the earth surface from 56°s outh to 60.25° north completely while the X-band data, distributed by the DLR (German Aerospace Center), cover it only partially. The C-band and Xband radar cannot penetrate the vegetation because of the short wavelength. Therefore, the height
Daily monitoring of marine mucilage using the MODIS products: a case study of 2021 mucilage bloom in the Sea of Marmara, Turkey
Environmental Monitoring and Assessment, Feb 11, 2022
Climate change and global warming along with human activities have caused abrupt changes in the a... more Climate change and global warming along with human activities have caused abrupt changes in the atmosphere, marine, and terrestrial ecosystems. One of these changes is the rising number of mucilage events in marine ecosystems. During the recent two decades, mucilage blooms have begun to appear more frequently in the Sea of Marmara in Turkey, surrounded by seven densely populated cities with various anthropogenic activities including household pollution, heavy industrialization, agricultural pollution, commercial fishing, shipyards, and specialized marine terminals with high maritime traffic density. In Spring 2021, a massive mucilage event observed in the Sea of Marmara plagued the entire ecosystem and raised awareness among the government and the public to detect and monitor this phenomenon. In this research, daily monitoring and detection of mucilage formations from the coarse spatial resolution MODIS products were investigated during the 2021 bloom period. The results were validated with the reference mucilage datasets derived from Sentinel-2A imagery and in situ spectroradiometer measurements over mucilage formations. The results revealed that the MODIS surface reflectance profiles were highly correlated with the field spectral measurements and estimated mucilage formations were spatially overlapped with Sentinel-2A. Overall, the produced maps accurately depicted the mucilage-covered areas despite the limitations of unreliable estimates along the land–water transition lines, and no-data areas due to the low-quality observations and high cloud coverage.
Photogrammetric Record, Oct 23, 2013
WorldView-2 (WV-2), whose panchromatic images have a 0Á5 m ground sampling distance (GSD), was la... more WorldView-2 (WV-2), whose panchromatic images have a 0Á5 m ground sampling distance (GSD), was launched by DigitalGlobe in 2009. It is the first commercial satellite to offer 8-band multispectral imagery with 1Á8 m resolution. Due to the offnadir sensor rotation of WV-2, it is feasible to obtain stereo coverage. Digital elevation models (DEMs) have been created with three WV-2 stereopairs of northern Istanbul with different land types and one of these is comprehensively analysed in this study. A reference DEM, developed from large-scale aerial photogrammetric mapping, together with a lidar DEM and an overlapping neighbouring WV-2 DEM, are used for validation. The generated WV-2 DEM reached, after filtering and in open areas, a standard deviation in height of approximately 1Á0 GSD. A higher number of discrepancies larger than 4 m exist than would be expected from a normal distribution, influencing the standard deviation more than the normalised median absolute deviation (NMAD).
Coherence Analysis of DSMS Generated by Multispectral RTK and RGB NON-RTK UAVS’ Simultaneous Data
IGARSS 2022 - 2022 IEEE International Geoscience and Remote Sensing Symposium, Jul 17, 2022
In the last decades, unmanned air vehicles (UAV) became one of the most demanded remote sensing t... more In the last decades, unmanned air vehicles (UAV) became one of the most demanded remote sensing technologies with the advantage of rapidly and periodically achievable low cost, high resolution and highly accurate data. Day by day, the technological level of the UAVs are improved by the integration of novel instruments. Major novel instruments are multispectral (MS) cameras and real-time kinematic global navigation satellite system (RTK GNSS) receivers. While MS cameras enable much better land cover classification performance, RTK GNSS receivers offer centimeter level UAV positioning accuracy without needing ground control points (GCP) and facilitate the field work before the flights. In this study, coherence of digital surface models (DSM) generated by MS RTK and RGB non-RTK UAVs' simultaneous data was investigated with modelbased comparison analysis in vegetation-dominant study area. During the analysis, GCP-supported and higher resolution non-RTK UAV DSM is appointed as the reference model. The results demonstrated that the horizontal coherence of the DSMs is around ±1 cm and the vertical coherence is around ±70 cm as standard deviation. The causes of vertical accuracy loss in MS RTK DSM are interpolation effect due to lower resolution point cloud and difficult high vegetation land cover, and the limited positioning potential of RTK GNSS.
Geomatik, Dec 1, 2021
Airborne laser scanning Forest stand height map DSM DTM NDSM Stand height information, a signific... more Airborne laser scanning Forest stand height map DSM DTM NDSM Stand height information, a significant parameter for forest inventory, is widely obtained with acquiring single tree information with geodetic techniques in underdeveloped or developing countries like Turkey. Acquired single tree information is generalized for whole forest stand with high error rate estimations. Developed countries, especially America and European countries, have completely excluded terrestrial methods with a high margin of error in forest stand height determination studies with the development of modern measurement technologies based on remote sensing (RS). This study aims to produce forest stand height map with airborne laser scanning (ALS) dense point clouds and to reveal the absolute location accuracy potential of the produced map. For this purpose, a forest-dominant study area was selected in Houston, USA, digital surface and terrain models were produced with ALS data, and a three-dimensional forest stand height map was obtained using a normalized digital surface model technique. The accuracy of the map produced was evaluated with model-based approaches using the forest stand height map produced with the data obtained from terrestrial laser scanning (TLS) technique in the same study area. International accuracy metrics for standard deviation and normalized median absolute deviation were used in the evaluations. In the light of the results, it has been determined that the stand height map produced with ALS data has an absolute position accuracy of ± 1 cm horizontally and ± 40 cm vertical.
International journal of environment and geoinformatics, Dec 15, 2021
Global warming threatens ecosystems through rising temperatures, increasing sea levels, drought, ... more Global warming threatens ecosystems through rising temperatures, increasing sea levels, drought, and extreme weather conditions. The natural balance of seas and oceans is also at stake with recent outbreaks of mucilage events all over the world. The mucilage phenomenon, which has been frequently observed in the Adriatic and Tyrrhenian seas, has taken place the second time in the Sea of Marmara in Spring 2021. The Sea of Marmara dividing the Asian and European parts of Turkey is an important inland sea with heavy maritime traffic, hosting many industrial zones and surrounded by highly populated cities. This study aims to determine the mucilage formations that were observed intensely all around the Sea of Marmara, focusing on the coasts of Istanbul, Kocaeli, Yalova, and Bursa through classifying Sentinel-2A images dated 19 and 24 May 2021, when the peak period of mucilage bloom, using a new paradigm of object-based image analysis (OBIA) approach. To create representative and homogenous image objects, multi-resolution segmentation was applied, and its result was inputted into a classification process using Random Forest (RF) classifier to generate thematic maps. The produced results were compared with pixel-based classification and a high correlation was estimated. Objectbased classification was found effective for the determination of mucilage-covered areas (> 90% overall accuracy) for both considered dates. More specifically, areas covered with mucilage aggregates were computed as 56.15 km² and 67.51 km² for 19 May and 24 May 2021, respectively, indicating rapid growth in only 5-day period. The resulting thematic maps revealed that mucilage was heavily distributed in the gulfs of Gemlik and Izmit and along the coasts of Darica, Tuzla and Pendik.
At the present time, remote sensing technologies are being improved and day by day, by launching ... more At the present time, remote sensing technologies are being improved and day by day, by launching new satellites with new developed techniques, images can be obtained faster with high resolution. The satellites which had been launched recent years for example; Ikonos, Quickbird and Orbview-3 have 1m and better resolution, continue their missions with quite success.
Optical Engineering, Dec 1, 2016
Airborne laser scanning (ALS) is one of the most effective remote sensing technologies providing ... more Airborne laser scanning (ALS) is one of the most effective remote sensing technologies providing precise three-dimensional (3-D) dense point clouds. A large-size ALS digital surface model (DSM) covering the whole Istanbul province was analyzed by point-based and model-based comprehensive statistical approaches. Point-based analysis was performed using checkpoints on flat areas. Model-based approaches were implemented in two steps as strip to strip comparing overlapping ALS DSMs individually in three subareas and comparing the merged ALS DSMs with terrestrial laser scanning (TLS) DSMs in four other subareas. In the model-based approach, the standard deviation of height and normalized median absolute deviation were used as the accuracy indicators combined with the dependency of terrain inclination. The results demonstrate that terrain roughness has a strong impact on the vertical accuracy of ALS DSMs. From the relative horizontal shifts determined and partially improved by merging the overlapping strips and comparison of the ALS, and the TLS, data were found not to be negligible. The analysis of ALS DSM in relation to TLS DSM allowed us to determine the characteristics of the DSM in detail.
Uploads
Papers by Umut Gunes Sefercik