Recognition of Building Roof Facets by Merging Aerial Images and 3D Lidar Data in a Hierarchical Segmentation Framework

2006 ◽  
Author(s):  
F. Bretar ◽  
M. Pierrot-Deseilligny ◽  
M. Roux
Keyword(s):  
Aerial Images ◽  
Lidar Data ◽  
Building Roof ◽  
Hierarchical Segmentation ◽  
Segmentation Framework ◽  
3D Lidar

scholarly journals BUILDING ROOF BOUNDARY EXTRACTION FROM LiDAR AND IMAGE DATA BASED ON MARKOV RANDOM FIELD

2017 ◽  
Vol XLII-1/W1 ◽  
pp. 339-344 ◽  
Author(s):  
A. P. Dal Poz ◽  
V. J. M. Fernandes
Keyword(s):  
Random Field ◽  
Markov Random Field ◽  
Image Data ◽  
Region Growing ◽  
Aerial Images ◽  
Aerial Image ◽  
Lidar Data ◽  
Building Roof ◽  
Markov Random ◽  
Bounding Boxes

In this paper a method for automatic extraction of building roof boundaries is proposed, which combines LiDAR data and highresolution aerial images. The proposed method is based on three steps. In the first step aboveground objects are extracted from LiDAR data. Initially a filtering algorithm is used to process the original LiDAR data for getting ground and non-ground points. Then, a region-growing procedure and the convex hull algorithm are sequentially used to extract polylines that represent aboveground objects from the non-ground point cloud. The second step consists in extracting corresponding LiDAR-derived aboveground objects from a high-resolution aerial image. In order to avoid searching for the interest objects over the whole image, the LiDAR-derived aboveground objects’ polylines are photogrammetrically projected onto the image space and rectangular bounding boxes (sub-images) that enclose projected polylines are generated. Each sub-image is processed for extracting the polyline that represents the interest aboveground object within the selected sub-image. Last step consists in identifying polylines that represent building roof boundaries. We use the Markov Random Field (MRF) model for modelling building roof characteristics and spatial configurations. Polylines that represent building roof boundaries are found by optimizing the resulting MRF energy function using the Genetic Algorithm. Experimental results are presented and discussed in this paper.

scholarly journals Building Change Detection Using Old Aerial Images and New LiDAR Data

2016 ◽  
Vol 8 (12) ◽  
pp. 1030 ◽  
Author(s):  
Shouji Du ◽  
Yunsheng Zhang ◽  
Rongjun Qin ◽  
Zhihua Yang ◽  
Zhengrong Zou ◽  
...  
Keyword(s):  
Change Detection ◽  
Aerial Images ◽  
Lidar Data

scholarly journals Multiscale Semantic Feature Optimization and Fusion Network for Building Extraction Using High-Resolution Aerial Images and LiDAR Data

2021 ◽  
Vol 13 (13) ◽  
pp. 2473
Author(s):  
Qinglie Yuan ◽  
Helmi Zulhaidi Mohd Shafri ◽  
Aidi Hizami Alias ◽  
Shaiful Jahari Hashim
Keyword(s):  
High Resolution ◽  
Large Scale ◽  
Spatial Information ◽  
Feature Fusion ◽  
Aerial Images ◽  
Semantic Gap ◽  
Superior Performance ◽  
Lidar Data ◽  
Building Extraction ◽  
Hierarchical Features

Automatic building extraction has been applied in many domains. It is also a challenging problem because of the complex scenes and multiscale. Deep learning algorithms, especially fully convolutional neural networks (FCNs), have shown robust feature extraction ability than traditional remote sensing data processing methods. However, hierarchical features from encoders with a fixed receptive field perform weak ability to obtain global semantic information. Local features in multiscale subregions cannot construct contextual interdependence and correlation, especially for large-scale building areas, which probably causes fragmentary extraction results due to intra-class feature variability. In addition, low-level features have accurate and fine-grained spatial information for tiny building structures but lack refinement and selection, and the semantic gap of across-level features is not conducive to feature fusion. To address the above problems, this paper proposes an FCN framework based on the residual network and provides the training pattern for multi-modal data combining the advantage of high-resolution aerial images and LiDAR data for building extraction. Two novel modules have been proposed for the optimization and integration of multiscale and across-level features. In particular, a multiscale context optimization module is designed to adaptively generate the feature representations for different subregions and effectively aggregate global context. A semantic guided spatial attention mechanism is introduced to refine shallow features and alleviate the semantic gap. Finally, hierarchical features are fused via the feature pyramid network. Compared with other state-of-the-art methods, experimental results demonstrate superior performance with 93.19 IoU, 97.56 OA on WHU datasets and 94.72 IoU, 97.84 OA on the Boston dataset, which shows that the proposed network can improve accuracy and achieve better performance for building extraction.

scholarly journals AN IMPROVED SNAKE MODEL FOR REFINEMENT OF LIDAR-DERIVED BUILDING ROOF CONTOURS USING AERIAL IMAGES

2016 ◽  
Vol XLI-B3 ◽  
pp. 583-589
Author(s):  
Qi Chen ◽  
Shugen Wang ◽  
Xiuguo Liu
Keyword(s):  
Urban Areas ◽  
Quality Index ◽  
Effective Means ◽  
Urban Air Pollution ◽  
High Elevation ◽  
Aerial Images ◽  
Snake Model ◽  
Building Roof ◽  
Searching Strategy ◽  
High Resolution Imagery

Building roof contours are considered as very important geometric data, which have been widely applied in many fields, including but not limited to urban planning, land investigation, change detection and military reconnaissance. Currently, the demand on building contours at a finer scale (especially in urban areas) has been raised in a growing number of studies such as urban environment quality assessment, urban sprawl monitoring and urban air pollution modelling. LiDAR is known as an effective means of acquiring 3D roof points with high elevation accuracy. However, the precision of the building contour obtained from LiDAR data is restricted by its relatively low scanning resolution. With the use of the texture information from high-resolution imagery, the precision can be improved. In this study, an improved snake model is proposed to refine the initial building contours extracted from LiDAR. First, an improved snake model is constructed with the constraints of the deviation angle, image gradient, and area. Then, the nodes of the contour are moved in a certain range to find the best optimized result using greedy algorithm. Considering both precision and efficiency, the candidate shift positions of the contour nodes are constrained, and the searching strategy for the candidate nodes is explicitly designed. The experiments on three datasets indicate that the proposed method for building contour refinement is effective and feasible. The average quality index is improved from 91.66% to 93.34%. The statistics of the evaluation results for every single building demonstrated that 77.0% of the total number of contours is updated with higher quality index.

Sign in / Sign up

Export Citation Format

Share Document