Dorota Iwaszczuk

Dorota Iwaszczuk
Technische Universität Darmstadt | TU · Department of Civil and Environmental Engineering Sciences (Dept.13)

Prof. Dr.-Ing.

About

39
Publications
5,145
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
158
Citations

Publications

Publications (39)
Article
Full-text available
Data are a key component for many applications and methods in the domain of photogrammetry and remote sensing. Especially data-driven approaches such as deep learning rely heavily on available annotated data. The amount of data is increasing significantly every day. However, reference data is not increasing at the same rate and finding relevant dat...
Conference Paper
Full-text available
The identification of Land Use and Land Cover is an important task for creating maps or monitoring surface changes. The results in this work were obtained using a deep neural network called VGG19-Unet. Influenced by the increasing success of separately implementing edge-extracted inputs in a model, this work used the Canny method to create such edg...
Article
Full-text available
Classification, and in particular semantic segmentation, plays a major role in remote sensing. In remote sensing, the classes usually correspond to landcover or landuse types while the data elements are image pixels. The results are so-called semantically segmented pixels describing the content of the data for each pixel. The identification of misc...
Article
Full-text available
Having a good estimate of the position and orientation of a mobile agent is essential for many application domains such as robotics, autonomous driving, and virtual and augmented reality. In particular, when using LiDAR and IMU sensors as the inputs, most existing methods still use classical filter-based fusion methods to achieve this task. In this...
Article
Full-text available
Indoor mapping has been gaining importance recently. One of the main applications of indoor maps is personal navigation. For this application, the connection to the outdoor map is very important, as users typically enter the building from outside and navigate to their destination inside. Obtaining this connection, however, is challenging, as the ge...
Conference Paper
Full-text available
Construction site planning is based on both explicit knowledge, as retrieved from regulations, and implicit knowledge, arising from experience. To retrieve and formalize rules from implicit knowledge, past construction projects can be analyzed. In this paper, we present an image analysis pipeline to retrieve information on past construction sites f...
Article
Full-text available
Indoor maps are required for multiple applications, such as navigation, building maintenance, and robotics. One of the common methods for map generation is laser scanning. In such maps, not only the geometry of the map is of interest, but also it's quality. This study aims at developing methods for real-time generation of indoor maps using features...
Article
Full-text available
In the last decade, we have observed an increasing demand for indoor scene modeling in various applications, such as mobility inside buildings, emergency and rescue operations, and maintenance. Automatically distinguishing between structural elements of buildings, such as walls, ceilings, floors, windows, doors etc., and typical objects in building...
Article
Full-text available
In this paper, we present our novel approach for the crowdsourced dynamic vertical mapping of buildings. For achieving this, we use the barometric sensor of smartphones to estimate altitude differences and the moment of the outdoor to indoor transition to extract reference pressure. We have identified the outdoor–indoor transition (OITransition) vi...
Article
Thermal infrared (TIR) images are often used to picture damaged and weak spots in the insulation of the building hull, which is widely used in thermal inspections of buildings. Such inspection in large-scale areas can be carried out by combining TIR imagery and 3D building models. This combination can be achieved via texture mapping. Automation of...
Conference Paper
Full-text available
In this paper, we present our developed and evaluated method for the dynamic mapping of the vertical characteristics inside a building. For achieving that, we extract data from smart-phone sensors and use those data for altitude estimation via the barometric formula. We introduce a novel approach for the extraction of reference pressure during th...
Article
Thermal properties of the building hull became an important topic of the last decade. Combining the thermal data with building models makes it possible to analyze thermal data in a 3D scene. In this paper we combine thermal images with 3D building models by texture mapping. We present a method for texture extraction from oblique airborne thermal in...
Article
Full-text available
Thermal properties of the building hull became an important topic of the last decade. Combining the thermal data with building models makes it possible to analyze thermal data in a 3D scene. In this paper we combine thermal images with 3D building models by texture mapping. We present a method for texture extraction from oblique airborne thermal in...
Article
This contribution is focused on the selection of building textures extracted from thermal infrared (TIR) image sequences acquired both from terrestrial and aerial platforms by introducing a quality assessment. Extracted quality features are completeness of the texture, projection accuracy, viewing angle, and geometric resolution. The calculation of...
Article
Full-text available
Thermal infrared imagery of urban areas became interesting for urban climate investigations and thermal building inspections. Using a flying platform such as UAV or a helicopter for the acquisition and combining the thermal data with the 3D building models via texturing delivers a valuable groundwork for large-area building inspections. However, su...
Article
Full-text available
This paper presents a method for identification of errors in 3D building models which are results of inaccurate creation process. Error detection is carried out within the camera pose estimation. As observations, parameters of the building corners and of the line segments detected in the image are used and conditions for the coplanarity of correspo...
Article
For image fusion in remote sensing applications the georeferencing accuracy using position, attitude, and camera calibration measurements can be insufficient. Thus, image processing techniques should be employed for precise coregistration of images. In this article a method for multimodal object-based image coregistration refinement between hypersp...
Article
Full-text available
ESPACE is an interdisciplinary Master's study programme of Technische Universitaet Muenchen (TUM) positioned at the interface between space technology and the engineering and natural science-based use of satellite data. It combines the technical aspects of the satellite and observation systems with scientific and commercial applications. A core top...
Article
Full-text available
3D city models are used in many fields. Photorealistic building textures find applications such as façade reconstruction, thermal building inspections and heat leakage detection using thermal infrared (TIR) images, quantitative evaluation or study of the materials lying on the object’s surface using multispectral images. Often texturing cannot be d...
Conference Paper
Full-text available
Data fusion techniques require a good registration of all the used datasets. In remote sensing, images are usually geo-referenced using the GPS and IMU data. However, if more precise registration is required, image processing techniques can be employed. We propose a method for multi-modal image coregistration between hyperspectral images (HSI) and...
Conference Paper
Full-text available
1 Zusammenfassung: Die Position und Orientierung eines Luftfahrzeuges im Raum wird durch GPS, oft mit einer IMU, bestimmt. Oft ist die mit GPS/IMU aufgenommene Position und Orientierung nicht genau genug um die in den aufgenommen Daten detektierten Objekte besser zu georeferenzieren und um die Daten aus verschiedenen Quellen zu fusionieren. In dies...
Article
Full-text available
Thermal building textures are used for the detection of damaged or weak spots in the insulation of building hulls. These textures can be extracted from directly geo-referenced oblique airborne infrared (IR) image sequences by projecting a 3D building model into the images. However, the direct geo-referencing is often not sufficiently accurate and t...
Conference Paper
Full-text available
Generation and texturing of building models is a fast developing field of research. Several techniques have been developed to extract building geometry and textures from multiple images and image sequences. In this paper, these techniques are discussed and extended to automatically add new textures from infrared (IR) image sequences to existing bui...
Conference Paper
Full-text available
Thermal building textures can be used for detection of damaged and weak spots in the building structure. These textures can be extracted from airborne infrared (IR) image sequences by projecting the 3D building model into the images. However, the direct georeferencing is often not sufficiently accurate and the projected 3D model does not match the...
Article
Full-text available
The aim of this article is to investigate methods for the automatic extraction of the infrared (IR) textures for the roofs and facades of existing building models. We focus on the correction of the measured exterior orientation parameters of the IR camera mounted on a mobile platform. The developed method is based on point-to-point matching of the...
Conference Paper
Thermal inspections of buildings contribute to detection of damaged and weak spots in the building hull. 3D spatial reference for this purpose can be achieved combining infrared images with 3D building models via texture mapping. Using terrestrial image sequences from a camera mounted in a mobile platform frontal faces can be captured, while airbor...
Conference Paper
Infrared (IR) images depict thermal radiation of physical objects. Imaging the building hull with an IR camera allows thermal inspections. Mapping these images as textures on 3D building models, 3D georeferencing of each pixel can be carried out. This is helpful for large area inspections. In IR images glass reflects the surrounding and shows false...
Conference Paper
Automatic texture mapping is an important task in enrichment of the common 3D city models. A significant part of all algorithms for automated texture mapping is the visibility checking. Nowadays most algorithms for texture extraction use visibility check based on z-buffer or polygon intersection in the image plane. Thus, the visibility of particula...
Article
Full-text available
Energy and climate changes are big topics in near future. In the European countries a significant part of consumed energy is used for heating in the buildings. Much effort is required for reducing this energy loss. Inspection and monitoring of buildings contribute in further development saving energy. For detection of areas with the highest loss of...

Network

Cited By

Projects

Projects (4)
Project
see description here: https://www.geodesy.tu-darmstadt.de/fernerkundung/forschung_fub/forschungsthemen_fub/fub_i_hm_fo_proj_dforest.en.jsp
Archived project
Thermal infrared (TIR) images are often used to picture damaged and weak spots in the insulation of the building hull, which is widely used in thermal inspections of buildings. The goal of this study is to provide a groundwork for such inspection in large-scale areas in the form of geo-referenced TIR textures for 3D building models. This avoids time consuming imaging and manually analyzing each face independently. It also enables the extraction of fa¸cade structures so that they can be embedded together with the TIR textures in a 3D spatial information system. For this purpose, 3D building models and TIR images need to be co-registered. Direct georeferencing as a co-registration method is usually not sufficient for precise texture extraction. Hence, an appropriate model-to-image matching is required. The majority of the existing solutions for model-to-image matching do not take the errors and uncertainties of the 3D models into account. Usually, textures are extracted for triangulated models and existing methods do not consider representation based on arbitrary polygons. Moreover, only few researchers assess the quality of extracted textures, but even they fail to pay attention to the quality of the fit between the 3D building models and the textures. Almost all the methods are designed for the visible domain and do not investigate other spectral bands. In this work, methods and strategies for precise texture extraction from airborne TIR image sequences are developed, and the potential for fa¸cade structures (such as windows) detection in the extracted textures is evaluated. In order to capture all faces, including the roofs, fa¸cades, and fa¸cades in the inner courtyard, an oblique looking video camera mounted on a flying platform is used. For this acquisition configuration, methods for a line-based model-to-image matching are developed, which consider uncertainties of the 3D building model, as well as of the image features, and determine the optimal exterior parameters of the camera. The remaining geometric mismatch between the projected 3D building model and image structures is compensated for every texture locally. This is done by adjusting the projected edges of the 3D building model to the gradient image generated from the TIR image. Moreover, this study investigates whether line tracking through the image sequence supports the matching.
Project
There are different ways to define 3D models, starting with the pure 3D point cloud, via derived polygon models with textures, to semantically labeled objects showing relations between them. Besides LiDAR, there are already commercial solutions for point cloud generation from images and high performance 3D information visualization, also for indoor environments. Since the interiors of 3D building models are typically generated separately, there are initiatives concentrating on their alignment with the outer 3D model. These methods focus on bringing outdoor and indoor geometries to one coordinate system, but they generally ignore semantics and topological relations between them that could be further exploited. Researchers are also looking for ways to retrieve semantic information form 3D data; such approaches, however, usually classify the outdoor and indoor geometries separately, without considering indoor-outdoor context information. The information about the relationship between outdoor and indoor spaces is especially useful to change the representation from surfaces to 3D objects, such as walls, roofs, windows, doors. Such representations are particularly important in design and construction, an industry that is becoming increasingly interested in using 3D models. 2D construction drawings are expected to be replaced by Building Information Modeling (BIM), where each construction element is a 3D object that can be manipulated separately (i.e., positioned, moved, changed, renovated.) Topological relationships between building elements are also important for simulations and building inspections, including thermal investigation, where outer and inner temperatures are connected. From the mobility point of view, transitional environments and their topology are of special interest, as the indoor-outdoor flow of people requires smooth transition. The goal of this research project is to perform a feasibility study by mapping an indoor environment in order to obtain its 3D geometry that will be seamlessly integrated with the outer geometry of a building, including the smooth integration of the semantic descriptions. In contrast to existing work on point cloud registration and semantic labeling, we focus on simultaneous labeling and relationship retrieval between the outdoor and indoor modeling. We primarily concentrate on three main objects classes, which will be detected in the 3D data: walls, windows, and doors. Windows are particularly important for the common registration of indoor and outdoor geometries and semantics as only they are directly observable in both datasets. Doors are also very important because they indicate topological relations between the rooms. Walls should be defined as volume, the space between two surfaces. The output of this project will be significantly enhanced 3D building models by providing high level of detail and semantic information, such as 3D building models in Level of Detail 4 (LOD4) or Building Information Models (BIM).