scholarly journals 3D RECONSTRUCTION OF BUILDINGS WITH GABLED AND HIPPED STRUCTURES USING LIDAR DATA

2014 ◽  
Vol XL-2/W3 ◽  
pp. 47-52
Author(s):  
H. Amini ◽  
P. Pahlavani ◽  
R. Karimi
Keyword(s):  
Urban Areas ◽  
3D Model ◽  
Remote Sensing Data ◽  
Training Data ◽  
Aerial Images ◽  
Lidar Data ◽  
Building Detection ◽  
3D Model Reconstruction ◽  
Accuracy And Precision ◽  
Intersection Line

Buildings are the most important objects in urban areas. Thus, building detection using photogrammetry and remote sensing data as well as 3D model of buildings are very useful for many applications such as mobile navigation, tourism, and disaster management. In this paper, an approach has been proposed for detecting buildings by LiDAR data and aerial images, as well as reconstructing 3D model of buildings. In this regard, firstly, building detection carried out by utilizing a Supper Vector Machine (SVM) as a supervise method. The supervise methods need training data that could be collected from some features. Hence, LiDAR data and aerial images were utilized to produce some features. The features were selected by considering their abilities for separating buildings from other objects. The evaluation results of building detection showed high accuracy and precision of the utilized approach. The detected buildings were labeled in order to reconstruct buildings, individually. Then the planes of each building were separated and adjacent planes were recognized to reduce the calculation volume and to increase the accuracy. Subsequently, the bottom planes of each building were detected in order to compute the corners of hipped roofs using intersection of three adjacent planes. Also, the corners of gabled roofs were computed by both calculating the intersection line of the adjacent planes and finding the intersection between the planes intersection line and their detected parcel. Finally, the coordinates of some nodes in building floor were computed and 3D model reconstruction was carried out. In order to evaluate the proposed method, 3D model of some buildings with different complexity level were generated. The evaluation results showed that the proposed method has reached credible performance.

Building Detection from Multiple Aerial Images in Dense Urban Areas

2001 ◽  
Vol 82 (3) ◽  
pp. 181-207 ◽  
Author(s):  
M. Fradkin ◽  
H. Maı̂tre ◽  
M. Roux
Keyword(s):  
Urban Areas ◽  
Aerial Images ◽  
Building Detection

scholarly journals A Hybrid GWR-Based Height Estimation Method for Building Detection in Urban Environments

2014 ◽  
Vol II-2 ◽  
pp. 23-29 ◽  
Author(s):  
X. Wei ◽  
X. Yao
Keyword(s):  
Urban Areas ◽  
Estimation Method ◽  
Urban Environments ◽  
Lidar Data ◽  
Building Detection ◽  
Building Roof ◽  
Aerial Photos ◽  
The Road ◽  
Urban Modelling ◽  
Gwr Model

LiDAR has become important data sources in urban modelling. Traditional methods of LiDAR data processing for building detection require high spatial resolution data and sophisticated methods. The aerial photos, on the other hand, provide continuous spectral information of buildings. But the segmentation of the aerial photos cannot distinguish between the road surfaces and the building roof. This paper develops a geographically weighted regression (GWR)-based method to identify buildings. The method integrates characteristics derived from the sparse LiDAR data and from aerial photos. In the GWR model, LiDAR data provide the height information of spatial objects which is the dependent variable, while the brightness values from multiple bands of the aerial photo serve as the independent variables. The proposed method can thus estimate the height at each pixel from values of its surrounding pixels with consideration of the distances between the pixels and similarities between their brightness values. Clusters of contiguous pixels with higher estimated height values distinguish themselves from surrounding roads or other surfaces. A case study is conducted to evaluate the performance of the proposed method. It is found that the accuracy of the proposed hybrid method is better than those by image classification of aerial photos along or by height extraction of LiDAR data alone. We argue that this simple and effective method can be very useful for automatic detection of buildings in urban areas.

scholarly journals LiDAR-Based System and Optical VHR Data for Building Detection and Mapping

Sensors ◽  
2020 ◽  
Vol 20 (5) ◽  
pp. 1285 ◽  
Author(s):  
Silvia Liberata Ullo ◽  
Chiara Zarro ◽  
Konrad Wojtowicz ◽  
Giuseppe Meoli ◽  
Mariano Focareta
Keyword(s):  
Environmental Monitoring ◽  
Case Studies ◽  
Urban Areas ◽  
Lidar Data ◽  
Building Detection ◽  
The Earth ◽  
Areas Of Interest ◽  
2D And 3D ◽  
3D Lidar ◽  
Very High

The aim of this paper is to highlight how the employment of Light Detection and Ranging (LiDAR) technique can enhance greatly the performance and reliability of many monitoring systems applied to the Earth Observation (EO) and Environmental Monitoring. A short presentation of LiDAR systems, underlying their peculiarities, is first given. References to some review papers are highlighted, as they can be regarded as useful guidelines for researchers interested in using LiDARs. Two case studies are then presented and discussed, based on the use of 2D and 3D LiDAR data. Some considerations are done on the performance achieved through the use of LiDAR data combined with data from other sources. The case studies show how the LiDAR-based systems, combined with optical Very High Resolution (VHR) data, succeed in improving the analysis and monitoring of specific areas of interest, specifically how LiDAR data help in exploring external environment and extracting building features from urban areas. Moreover the discussed Case Studies demonstrate that the use of the LiDAR data, even with a low density of points, allows the development of an automatic procedure for accurate building features extraction, through object-oriented classification techniques, therefore by underlying the importance that even simple LiDAR-based systems play in EO and Environmental Monitoring.

scholarly journals Multi-Object Segmentation in Complex Urban Scenes from High-Resolution Remote Sensing Data

2021 ◽  
Vol 13 (18) ◽  
pp. 3710
Author(s):  
Abolfazl Abdollahi ◽  
Biswajeet Pradhan ◽  
Nagesh Shukla ◽  
Subrata Chakraborty ◽  
Abdullah Alamri
Keyword(s):  
Remote Sensing ◽  
High Resolution ◽  
Urban Areas ◽  
Object Segmentation ◽  
Remote Sensing Data ◽  
Semantic Segmentation ◽  
Aerial Images ◽  
Urban Scenes ◽  
Sensing Data ◽  
Boundary Information

Terrestrial features extraction, such as roads and buildings from aerial images using an automatic system, has many usages in an extensive range of fields, including disaster management, change detection, land cover assessment, and urban planning. This task is commonly tough because of complex scenes, such as urban scenes, where buildings and road objects are surrounded by shadows, vehicles, trees, etc., which appear in heterogeneous forms with lower inter-class and higher intra-class contrasts. Moreover, such extraction is time-consuming and expensive to perform by human specialists manually. Deep convolutional models have displayed considerable performance for feature segmentation from remote sensing data in the recent years. However, for the large and continuous area of obstructions, most of these techniques still cannot detect road and building well. Hence, this work’s principal goal is to introduce two novel deep convolutional models based on UNet family for multi-object segmentation, such as roads and buildings from aerial imagery. We focused on buildings and road networks because these objects constitute a huge part of the urban areas. The presented models are called multi-level context gating UNet (MCG-UNet) and bi-directional ConvLSTM UNet model (BCL-UNet). The proposed methods have the same advantages as the UNet model, the mechanism of densely connected convolutions, bi-directional ConvLSTM, and squeeze and excitation module to produce the segmentation maps with a high resolution and maintain the boundary information even under complicated backgrounds. Additionally, we implemented a basic efficient loss function called boundary-aware loss (BAL) that allowed a network to concentrate on hard semantic segmentation regions, such as overlapping areas, small objects, sophisticated objects, and boundaries of objects, and produce high-quality segmentation maps. The presented networks were tested on the Massachusetts building and road datasets. The MCG-UNet improved the average F1 accuracy by 1.85%, and 1.19% and 6.67% and 5.11% compared with UNet and BCL-UNet for road and building extraction, respectively. Additionally, the presented MCG-UNet and BCL-UNet networks were compared with other state-of-the-art deep learning-based networks, and the results proved the superiority of the networks in multi-object segmentation tasks.

Sign in / Sign up

Export Citation Format

Share Document