Content uploaded by Umut Gunes Sefercik
Author content
All content in this area was uploaded by Umut Gunes Sefercik on Jan 24, 2021
Content may be subject to copyright.
A preview of the PDF is not available
SFM Tabanlı Yeni Nesil Görüntü Eşleştirme Yazılımlarının Fotogrametrik 3B Modelleme Potansiyellerinin Karşılaştırması
Abstract
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. Keywords ABSTRACT SFM VisualSFM Agisoft Pix4D Digital Surface Model Structure from motion (SFM) matching algorithm is the basic principle of new generation and widely used image matching software. Although these software work in common principle, their final products may contain different characteristics and distortions depending on their buried parameters. In the literature, there is lack of publishments which compare the three dimensional modelling performance of SFM based new generation software. Accordingly, our research group decided to carry out a study that could be a reference for upcoming researches. In this study, using VisualSFM, Agisoft and Pix4D SFM based image matching software, 3D digital surface models (DSM) were generated from unmanned air vehicle (UAV) high resolution aerial photos in a Campus of Zonguldak Bulent Ecevit University. Generated DSMs were comprehensively evaluated and compared by visual and statistical approaches utilizing the Agisoft DSM as the reference. The results clearly demonstrated the pros and cons of each analyzed SFM-based software.
... Structure From Motion (SFM) iki boyutlu görüntülerin birleştirilmesinden 3B yapıları oluşturabilmek için geliştirilen algoritmadır. Fotogrametrik görüntü eşleştirme işlemi yapan bu algoritma yaygın olarak kullanılmakta olan yeni nesil bulut tabanlı görüntü eşleme yazılımlarının genel işlevidir [4]. Sefercik [5]. ...
... Optical UAVs have difficulties in autonomous navigation only in regions where the GNSS equipment they use can be blocked (military areas, etc.), and this situation can be overcome if users do not fly close to the relevant regions (Lu et al. 2018). Users can obtain three-dimensional (3D) models and orthomosaics by easily processing aerial photographs obtained from optical UAVs with the help of easy and user-friendly interfaces in SfM (structure from motion) based package softwares (Sefercik et al. 2020). However, as with all remote sensing systems, the visual quality of products produced by UAV technology can often be misleading. ...
The quality of the final products produced with unmanned aerial vehicle (UAV) photogrammetry is increasing day by day; along with this, the process steps in the photogrammetric process are detailed, and work on quality improvement of products continues. Camera optimization is an important step in the processing of images in digital photogrammetry. Camera alignment optimization is an optimization process before dense point cloud generation with the help of a sparse point cloud generated after reciprocal orientation, which is the preliminary stage of creating a viable substantive model based on the dense point cloud. This process is recommended to users in package software in order to increase geometric accuracy and in light of the findings on absolute orientation, its impact is understood. In this work, the impact of camera optimization on the quality of three-
dimensional (3D) models created from RGB UAV aerial photos collected with polygonal flights was examined. The data of the study were collected with a polygonal flight from a height of 120m with a DJI Phantom IV Pro V2 RGB UAV. With the help of these data, digital surface models were produced separately with and without optimization. In order to compare all of the pixels models contain, model-based methodologies were used for 3D compatibility analyses. Compatibility analyzes were carried out for the building, open land, and forest classes in the land, and in addition, the analysis of the effect of land slope on the results was also examined. Interior orientation can be automatically resolved with the metadata of the UAV cameras, therefore, there is no need to interfere with the parameters during the optimization process. In this study, radial distortion (k4) and skew (b1, b2) parameters were included in addition to the parameters presented by default in optimization, and when the results were examined, an improvement in absolute orientation was observed, while serious deterioration in model accuracy was detected. This disruptive effect was statistically explained in detail and the necessity of the optimization was highlighted.
... The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLVI-4/W5-2021 The 6th International Conference on Smart City Applications, 27-29 October 2021, Karabuk University, Virtual Safranbolu, Turkey stereoscopic photogrammetry (Dereli et al., 2019;Sefercik et al., 2020). It uses overlapping photos from multiple viewpoints to reconstruct camera position and camera geometry. ...
Unmanned air vehicle (UAV) became an alternative airborne remote sensing technique, due to providing very high resolution and low cost spatial data and short processing time. Particularly, optical UAVs are frequently utilized in various applications such as mapping, agriculture, and forestry. Especially for precise agriculture purposes, the UAVs were equipped with multispectral cameras which enables to classify land cover easily. In this study, the land cover classification potential of DJI Phantom IV Multispectral, one of the most preferred agricultural UAVs in the world, was investigated using spectral angle mapper, minimum distance and maximum likelihood pixel-based classification techniques and object-based classification. In the investigation, a part of Gebze Technical University (GTU) Northern Campus, includes a large variety of land cover classes, was selected as the study area. The UAV aerial photos were achieved from 70 m flight altitude and processed using structure from motion (SfM)-based image matching software Agisoft Metashape. The pixel-based and object-based land cover classification processes were completed with ENVI and eCognition software respectively. 16 independent land cover classes were classified and the results demonstrated that the accuracies are 73.46% in spectral angle mapper, 75.27% in minimum distance and 93.56% in maximum likelihood pixel-based classification techniques and 90.09% in nearest neighbour object-based classification.
... As the commercial software market has expanded in recent years, many software packages using the sfm algorithm have emerged. In this case, it has been a matter of debate which of the several photogrammetric software packages available on point clouds of the underwater environment can provide better results (Drap et al., 2015;Çelik et al., 2020;Sefercik et al., 2020;Ulvi et al., 2020;Yiğit et al. Uivi, 2020). ...
... Orientation of aerial photos and dense point cloud generation were done using Agisoft Metashape photogrammetric evaluation software. Agisoft Metashape uses the structure from motion (SfM) technique which is based on the principle of stereoscopic photogrammetry (Dereli et al., 2019;Sefercik et al., 2020). Using a SFM based software, camera position and 3D geometry can be reconstructed by utilizing overlapping photos that were taken from different viewpoints. ...
Unmanned air vehicle (UAV) has become an indispensable mobile mapping technology of remote sensing thanks to offering low cost and high resolution spatial data. Particularly, camera equipped optical UAVs are large in demand by land-related professions, including mapping, agriculture and forestry. Regarding the requirements, the technological level of the optical UAVs rises day by day by adding novel payloads. For instance, global navigation satellite system (GNSS) receivers with real-time kinematic (RTK) positioning capability were added to facilitate the fieldwork for ground control point (GCP) set up and measurements before UAV flights. Multispectral cameras were added to increase the automatic land cover classification potential of generated ortho-mosaics. At this point, the most significant question is the contribution level of these technological payloads. In this study, our research group evaluated the RTK GNSS positioning accuracy and automatic land cover classification potential of "DJI Phantom IV Multispectral RTK", which is one of the most common optical UAVs for scientific and commercial applications. For the evaluations, a study area that includes a large variety of land cover classes was selected. The UAV RTK GNSS positioning accuracy was calculated by comparing the RTK GNSS data obtained from the UAV with the measured GCPs in the study area. Furthermore, the land cover classification performance of Multispectral UAV was analysed by pixel and object-based classification techniques separately. For this purpose, while spectral angle mapper (SAM), minimum distance (MD) and maximum likelihood (ML) classifiers were applied to perform pixel-based classification, nearest neighbour (NN) classifier was employed to utilize object-based classification. The positioning accuracy results demonstrated that the root mean square error (RMSE) of UAV RTK GNSS is ±1.1 cm in X, ±2.7 cm in Y, and ±5.7 cm in Z. The classification results showed that the highest overall accuracy was estimated as 93.56% with ML classifier and its classification performance was found to be superior compared to those of SAM (73.46%) and MD (75.27%) classifiers. On the other hand, the overall accuracy was calculated as 90.09% for object-based classification and it was 3% lower than the pixel-based ML classification result. This could be the result of heterogeneity of the image objects created during the segmentation stage. Further studies are required to improve the object-based classification accuracy by applying different segmentation methods and quality measures.
Su altındaki bir cismin üç boyutlu olarak belgelenmesi ve dijital platformlara aktarılması son yıllarda önem kazanmıştır. Bu amaçla fotogrametri yöntemi su altında denenmiş ve sualtı fotogrametrisi terimi literatürde kendine yer bulmuştur. Fotogrametri yönteminin tercih edilmesinin en önemli nedeni geleneksel yöntemlere göre daha kısa sürede veri üretimi, zaman ve maliyet açısından olumlu katkı sağlamasıdır. Ancak su altında fotografik veri toplayan dalgıçlar su altında sınırlı bir süre kalabildikleri için fotoğraflama işlemi tekrarlanmakta ve süreç uzamaktadır. Bu olumsuzluğu en aza indirmek için video çekim yöntemiyle veri elde etme çalışmaları denenmeye başlanmıştır. Bu çalışmada, bir havuza test amaçlı obje yerleştirilerek hem fotoğraf çekim hem de video çekim yöntemi ile üç boyutlu model üretiminin karşılaştırılmalı analizi yapılmıştır. Çalışma sonucunda video çekim yönteminin doğruluğunu test etmek için fotoğraf çekim yöntemi referans kabul edilmiştir. Her iki veriden elde edilen üç boyutlu nokta bulutları karşılaştırılarak doğruluk analizi yapılmış video çekim yönteminin karesel ortalama hatası ± 3.24 cm olarak tespit edilmiştir. Bu çalışma ile su altında video çekim yönteminin kullanılabilirliği araştırılmış sonuç olarak video çekim yönteminin doğruluk açısında yeterli düzeyde olduğu tespit edilmiş fakat görsel açıdan yetersiz bulunmuştur.
İnsansız hava aracı (İHA) teknolojisi sonuç ürünlerinin en doğru geometride elde edilebilmesi ve ileri üretimlerde kullanılabilmesi için fotogrametrik işlem adımlarında pek çok parametrenin ve standartın doğru uygulanması gerekmektedir. Bu standartların başında gelenlerden biri de görüntü alımı yapan kameraların optimizasyonudur. İHA fotogrametrisinde kamera optimizasyonu, yoğun nokta bulutuna dayalı uygulanabilir asli bir model oluşturmanın ön aşaması olan ve karşılıklı yöneltme sonrası ilk olarak üretilen seyrek nokta bulutundaki az sayıda noktaya göre İHA kamerasının yeniden hizalanmasıdır. Bu işlem, geometrik yöneltme doğruluğunu artırmak için paket yazılımlarda kullanıcılar tarafından sıkça tercih edilmekte ve etkisi mutlak yöneltme sonuçlarına göre yorumlanmaktadır. Bu çalışmada, poligonal uçuşlarla elde edilen RGB İHA hava fotoğraflarından üretilen üç boyutlu (3B) modellerin kaliteleri üzerinde kamera optimizasyonunun etkileri araştırılmıştır. Araştırmada, 120 m irtifada yapılan DJI Phantom IV V2 RGB İHA uçuşlarından elde edilen veriler kullanılarak kamera optimizasyonu uygulanmış ve uygulanmamış şekilde ayrı ayrı dijital yüzey modelleri (DYM) üretilmiş ve içerdikleri tüm piksellerin karşılaştırılmasına olanak veren model bazlı yaklaşımlar ile 3B uyum analizleri gerçekleştirilmiştir. Analizler, açık, bina ve orman sınıflarında ayrı ayrı yapılmış, ek olarak arazi eğiminin sonuçlar üzerindeki etkisi de irdelenmiştir. İHA dijital kameraları, iç yöneltme parametrelerinin doğrudan hava fotoğrafı metaverilerinden çözülmesine olanak verdiğinden optimizasyonda müdahale gerektirmemektedir. Optimizasyonda kullanılan iç yöneltme dışındaki parametrelerin kullanımında ise kullanıcı çok dikkatli olmalıdır. Bu araştırmada, program tarafından sunulan tüm parametrelerin eş zamanlı değerlendirilmesinin model üzerinde ki etkilerinin test edilmesi amacıyla kamera optimizasyonunda varsayılan olarak sunulan iç yöneltme ve kamera distorsiyon katsayılarına ek radyal distorsiyon (k4) ve eğiklik (b1,b2) parametreleri de işlem sürecine dahil edilmiştir. Analizler sonucunda, tercih edilen değişkenlere de bağlı olarak, optimizasyonun mutlak yöneltme hassasiyetini arttırırken model doğruluğu üzerinde bilimsel araştırma ihtiyacı doğuracak boyutta negatif etkileri olduğu görülmüştür. Bu bozucu etkiler göz önünde bulundurulduğunda, kullanıcı tarafından tüm işlem adımlarında kontrollü bir ilerleyiş veya optimizasyonun gerekliliği hususunda analiz ihtiyacının zorunluluğu görsel ve istatistiki verilerle ortaya konulmuştur.
Bu çalışmada, sulama suyu amaçlı kullanılan Aydınlar (Gülüç) çayının fiziko-kimyasal parametreler açısından su kalitesi incelenmiş ve insansız hava aracı (İHA) verilerinin spektral analizi ile karşılaştırmalı değerlendirmeleri yapılmıştır. Çalışma kapsamında, su kalitesini belirlemek için üç farklı numune alım noktası belirlenmiş ve Haziran 2019 (kurak dönem), Eylül 2019 ve Aralık 2019 (yağışlı dönem) tarihlerinde çayı temsilen su numuneleri alınmıştır. Alınan numunelerin pH, elektriksel iletkenlik (Eİ), bulanıklık, askıda katı madde (AKM), sıcaklık değerleri ile kimyasal oksijen ihtiyacı (KOİ), biyokimyasal oksijen ihtiyacı (BOİ5), çözünmüş oksijen (ÇO), toplam kjeldahl azotu (TKA), toplam fosfat (TP) ve yağ-gres analizleri gerçekleştirilmiştir. pH, Eİ, bulanıklık, sıcaklık ve AKM değerleri
sırasıyla ortalama 7.67-8.32, 815-1106 µS/cm, 14.91-20.95 NTU, 9.34-25.34 °C ve 32.84-82.67 mg/L arasında, ÇO, BOİ, KOİ, TP, TKA ve yağ-gres değerleri ise sırasıyla ortalama 2.34-3.40 mg/L, 8.14-20.14 mg/L, 96.44-132.7 mg/L, 0.179-0.296 mg/L, 13.91-20.14
mg/L ve 0.0087-0.016 mg/L arasında ölçülmüştür. Tespit edilen fiziko-kimyasal parametreler Yerüstü Su Kalitesi Yönetmeliği (YSKY)’ne göre değerlendirilmiştir. Ayrıca, belirlenen fiziko-kimyasal parametrelerin su yüzeyinde meydana getirdiği spektral (renk)
değişimler, yüksek çözünürlüklü kamera donanımına sahip İHA ile elde edilen hava fotoğraflarının işlenmesi sonucu üretilen ortomozaikler üzerinden analiz edilmiştir. Sonuç olarak, Aydınlar Çay’ının ÇO, BOİ5, KOİ, TP ve TKA ortalama konsantrasyonları
YSKY’in de belirtilen değerlere göre III. Sınıf ve IV. Sınıf su kalite özelliği gösterdiği tespit edilmiştir.
Nowadays, Unmanned Aerial System (UAS)-based photogrammetry offers an affordable, fast and effective approach to real-time acquisition of high resolution geospatial information and automatic 3D modelling of objects for numerous applications such as topography mapping, 3D city modelling, orthophoto generation, and cultural heritages preservation. In this paper, the capability of four different state-of-the-art software packages as 3DSurvey, Agisoft Photoscan, Pix4Dmapper Pro and SURE is examined to generate high density point cloud as well as a Digital Surface Model (DSM) over a historical site. The main steps of this study are including: image acquisition, point cloud generation, and accuracy assessment. The overlapping images are first captured using a quadcopter and next are processed by different software to generate point clouds and DSMs. In order to evaluate the accuracy and quality of point clouds and DSMs, both visual and geometric assessments are carry out and the comparison results are reported.
Structure from Motion with Multi View Stereo provides hyperscale landform models using images acquired from standard compact cameras and a network of ground control points. The technique is not limited in temporal frequency and can provide point cloud data comparable in density and accuracy to those generated by terrestrial and airborne laser scanning at a fraction of the cost. It therefore offers exciting opportunities to characterise surface topography in unprecedented detail and, with multi-temporal data, to detect elevation, position and volumetric changes that are symptomatic of earth surface processes. This book firstly places Structure from Motion in the context of other digital surveying methods and details the Structure from Motion workflow including available software packages and assessments of uncertainty and accuracy. It then critically reviews current usage of Structure from Motion in the geosciences, provides a synthesis of recent validation studies and looks to the future by highlighting opportunities arising from developments in allied disciplines. This book will appeal to academics, students and industry professionals because it balances technical knowledge of the Structure from Motion workflow with practical guidelines for image acquisition, image processing and data quality assessment and includes case studies that have been contributed by experts from around the world.
Single photon lidar (SPL) is an innovative technology for rapid forest structure and terrain characterization over large areas. Here, we evaluate data from an SPL instrument - the High Resolution Quantum Lidar System (HRQLS) that was used to map the entirety of Garrett County in Maryland, USA (1700 km2). We develop novel approaches to filter solar noise to enable the derivation of forest canopy structure and ground elevation from SPL point clouds. SPL attributes are compared with field measurements and an existing leaf-off, low-point density discrete return lidar dataset as a means of validation. We find that canopy and ground characteristics from SPL are similar to discrete return lidar despite differences in wavelength and acquisition periods but the higher point density of the SPL data provides more structural detail. Our experience suggests that automated noise removal may be challenging, particularly over high albedo surfaces and rigorous instrument calibration is required to reduce ground measurement biases to accepted mapping standards. Nonetheless, its efficiency of data collection, and its ability to produce fine-scale, three-dimensional structure over large areas quickly strongly suggests that SPL should be considered as an efficient and potentially cost-effective alternative to existing lidar systems for large area mapping.
Stable imaging of an unmanned aerial vehicle (UAV) photogrammetry system is an important issue that affects the data processing and application of the system. Compared with traditional aerial images, the large rotation of roll, pitch, and yaw angles of UAV images decrease image quality and result in image deformation, thereby affecting the ground resolution, overlaps, and the consistency of the stereo models. These factors also cause difficulties in automatic tie point matching, image orientation, and accuracy of aerial triangulation (AT). The issues of large-angle photography of UAV photogrammetry system are discussed and analyzed quantitatively in this paper, and a simple and lightweight three-axis stabilization platform that works with a low-precision integrated inertial navigation system and a three-axis mechanical platform is used to reduce this problem. An experiment was carried out with an airship as the flight platform. Another experimental dataset, which was acquired by the same flight platform without a stabilization platform, was utilized for a comparative test. Experimental results show that the system can effectively isolate the swing of the flying platform. To ensure objective and reliable results, another group of experimental datasets, which were acquired using a fixed-wing UAV platform, was also analyzed. Statistical results of the experimental datasets confirm that stable imaging of a UAV platform can help improve the quality of aerial photography imagery and the accuracy of AT, and potentially improve the application of images acquired by a UAV.
In this paper，the aforementioned three major aspects related to the Unmanned Aerial Vehicles (UAV) system for low altitude aerial photogrammetry, i.e., flying platform, imaging sensor system and data processing software, are discussed. First of all, according to the technical requirements about the least cruising speed, the shortest taxiing distance, the level of the flight control and the performance of turbulence flying, the performance and suitability of the available UAV platforms (e.g., fixed wing UAVs, the unmanned helicopters and the unmanned airships) are compared and analyzed. Secondly, considering the restrictions on the load weight of a platform and the resolution pertaining to a sensor, together with the exposure equation and the theory of optical information, the principles of designing self-calibration and self-stabilizing combined wide-angle digital cameras (e.g., double-combined camera and four-combined camera) are placed more emphasis on. Finally, a software named MAP-AT, considering the specialty of UAV platforms and sensors, is developed and introduced. Apart from the common functions of aerial image processing, MAP-AT puts more effort on automatic extraction, automatic checking and artificial aided adding of the tie points for images with big tilt angles. Based on the recommended process for low altitude photogrammetry with UAVs in this paper, more than ten aerial photogrammetry missions have been accomplished, the accuracies of Aerial Triangulation, Digital orthophotos(DOM)and Digital Line Graphs(DLG) of which meet the standard requirement of 1:2000, 1:1000 and 1:500 mapping.
Image matching is a key procedure in the process of generation of Digital Surface Models (DSM). We have developed a new approach for image matching and the related software package. This technique has proved its good performance in many applications. Here, we demonstrate its use in 3D tree modelling. After a brief description of our image matching technique, we show results from analogue and digital aerial images and high-resolution satellite images (IKONOS). In some cases, comparisons with manual measurements and/or airborne laser data have been performed. The evaluation of the results, qualitative and quantitative, indicate the very good performance of our matcher. Depending on the data acquisition parameters, the photogrammetric DSM can be denser than a DSM generated by laser, and its accuracy may be better than that from laser, as in these investigations. The tree canopy is well modelled, without smoothing of small details and avoiding the canopy penetration occurring with laser. Depending on the image scale, not only dense forest areas but also individual trees can be modelled.
The objective of this investigation was to develop and investigate methods for point cloud generation by image matching using aerial image data collected by quadrocopter type micro unmanned aerial vehicle (UAV) imaging systems. Automatic generation of high-quality, dense point clouds from digital images by image matching is a recent, cutting-edge step forward in digital photogrammetric technology. The major components of the system for point cloud generation are a UAV imaging system, an image data collection process using high image overlaps, and post-processing with image orientation and point cloud generation. Two post-processing approaches were developed: one of the methods is based on Bae Systems' SOCET SET classical commercial photogrammetric software and another is built using Microsoft(®)'s Photosynth™ service available in the Internet. Empirical testing was carried out in two test areas. Photosynth processing showed that it is possible to orient the images and generate point clouds fully automatically without any a priori orientation information or interactive work. The photogrammetric processing line provided dense and accurate point clouds that followed the theoretical principles of photogrammetry, but also some artifacts were detected. The point clouds from the Photosynth processing were sparser and noisier, which is to a large extent due to the fact that the method is not optimized for dense point cloud generation. Careful photogrammetric processing with self-calibration is required to achieve the highest accuracy. Our results demonstrate the high performance potential of the approach and that with rigorous processing it is possible to reach results that are consistent with theory. We also point out several further research topics. Based on theoretical and empirical results, we give recommendations for properties of imaging sensor, data collection and processing of UAV image data to ensure accurate point cloud generation.
The extraction of the third dimension from remote sensing data is a well known technique. Since in a number of countries aerial images and laser scanner data are unavailable, expensive or classified, stereoscopic high-resolution optical satellite images provide a viable alternative
for generating digital surface and digital terrain models. Especially the automatic extraction of highly accurate 3D surface models in urban areas is still a very complicated task due to occlusions, large differences in height and the variety of objects and surface material.
In this
paper an analysis and a visual and quantitative comparison of three different matching algorithms for generating urban DSMs based on very high-resolution satellite images is presented. The three algorithms are least squares matching (LSM) in a region growing fashion, dynamic programming (DP)
and semiglobal matching (SGM). The characteristics of the three algorithms as applied to four different Ikonos stereo pairs with a ground sampling distance of 1 m are shown. The following results were obtained: visually, in the LSM results the shape of the buildings is considerably smoothed.
While in the DP results the building shape is sharper, only little detail is visible on the building roofs, and streaking along the epipolar lines causes problems. With SGM more details can be extracted and the results visually have the best quality. Based on reference data for the different
test sites, the standard deviation of the building heights determined by LSM and DP is in the range of one pixel or slightly better, while it is in the range of half a pixel for SGM.