Building Change Detection Based on Deep Learning and Belief Function

Yuan, Xiangtian; Tian, Jiaojiao; Reinartz, Peter

doi:10.1109/jurse.2019.8808968

Cited by 6 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, several studies have introduced multimodal strategies for change detection. For example, in our previous works the decision fusion method belief functions have been proven to be an efficient fusion module for multimodal change detection [60], [83], [84], which can effectively improve the building change detection results compared with single-modal change indicators. The paper [83] proposes a change detection pipeline based on the robust height differences between DSMs and the similarity measurement between corresponding optical image pairs.…”

Section: B Change Detection With Multimodal Datamentioning

confidence: 99%

“…Finally, four decision-making criteria are employed to convert the fused global BBAs to building change maps. [60] extends the framework in [84] and employs initial building probabilities extracted by the deep neural network Deeplabv3+ for the change decision, which shows better generalization ability than the previous version. Also based on the Dempster-Shafer theory, [85] introduces a complementary evidence fusion framework.…”

Section: B Change Detection With Multimodal Datamentioning

confidence: 99%

“…Nevertheless, end-to-end deep learning-based multimodal change detection methods have not been widely investigated, which is partly due to the lack of sufficient public datasets [72], [87]. Although [60] involves deep learning, it only uses the network for building extraction rather than change detection. The lack of sufficient multimodal change detection data impedes the development of robust end-to-end methods with strong cross-domain generalizability.…”

Section: B Change Detection With Multimodal Datamentioning

confidence: 99%

“…Desiring beyond homogeneous data, a few learning-based studies have shifted the focus from singlemodal methods to multimodal data fusion, enriching the features or probabilities via a fusion operation (e.g., summation, average, concatenation, etc.). 2D-2.5D/3D data fusion utilizing multimodal data as inputs for a fusion framework may increase the accuracy of change detection [60]. Recently, multimodal knowledge transfer semi-supervised learning architecture represented by co-learning utilizing multimodal data pairs only for the training phase [61], [62] has attracted the attention in remote sensing tasks such as building extraction [51], [58] and semantic segmentation [63].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Multimodal Co-Learning for Building Change Detection: A Domain Adaptation Framework Using VHR Images and Digital Surface Models

Xie,

Yuan,

Zhu

et al. 2024

IEEE Trans. Geosci. Remote Sensing

Self Cite

View full text Add to dashboard Cite

In this article, we propose a multimodal co-learning framework for building change detection. This framework can be adopted to jointly train a Siamese bitemporal image network and a height difference map (HDiff) network with labeled source data and unlabeled target data pairs. Three co-learning combinations (vanilla co-learning, fusion co-learning, and detached fusion colearning) are proposed and investigated with two types of colearning loss functions within our framework. Our experimental results demonstrate that the proposed methods are able to take advantage of unlabeled target data pairs and therefore enhance the performance of single-modal neural networks on the target data. In addition, our synthetic-to-real experiments demonstrate that the recently published synthetic dataset SMARS is feasible to be used in real change detection scenarios, where the optimal result is with the F1 score of 79.29%.

show abstract

Section: B Change Detection With Multimodal Datamentioning

confidence: 99%

Section: B Change Detection With Multimodal Datamentioning

confidence: 99%

Section: B Change Detection With Multimodal Datamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Multimodal Co-Learning for Building Change Detection: A Domain Adaptation Framework Using VHR Images and Digital Surface Models

Xie,

Yuan,

Zhu

et al. 2024

IEEE Trans. Geosci. Remote Sensing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Tian and Qin et al [30,31] used the random forest to extract building probability map and then combined multi-temporal building probability map with the 3D bilateral filter. Yuan et al [32] extracted the building probability map using deep learning and integrated it with the 3D change feature based on DS fusion theory. Pang et al [33] used a graph-cuts-based algorithm to extract changed objects from two-temporal DSMs and used the corresponding aerial images to remove non-building objects with a structural feature.…”

Section: Introductionmentioning

confidence: 99%

Building Change Detection Based on 3D Co-Segmentation Using Satellite Stereo Imagery

Wang

Zhang

et al. 2022

Remote Sensing

View full text Add to dashboard Cite

Building change detection using remote sensing images is significant to urban planning and city monitoring. The height information extracted from very high resolution (VHR) satellite stereo images provides valuable information for the detection of 3D changes in urban buildings. However, most existing 3D change detection algorithms are based on the independent segmentation of two-temporal images and the feature fusion of spectral change and height change. These methods do not consider 3D change information and spatial context information simultaneously. In this paper, we propose a novel building change detection algorithm based on 3D Co-segmentation, which makes full use of the 3D change information contained in the stereoscope data. An energy function containing spectral change information, height change information, and spatial context information is constructed. Image change feature is extracted using morphological building index (MBI), and height change feature is obtained by robust normalized digital surface models (nDSM) difference. 3D Co-segmentation divides the two-temporal images into the changed foreground and unchanged background through the graph-cut-based energy minimization method. The object-to-object detection results are obtained through overlay analysis, and the quantitative height change values are calculated according to this correspondence. The superiority of the proposed algorithm is that it can obtain the changes of buildings in planar and vertical simultaneously. The performance of the algorithm is evaluated in detail using six groups of satellite datasets. The experimental results prove the effectiveness of the proposed building change detection algorithm.

show abstract

Rapid identification of damaged buildings using incremental learning with transferred data from historical natural disaster cases

Tang

Yang

et al. 2023

ISPRS Journal of Photogrammetry and Remote Sensing

View full text Add to dashboard Cite

Building Change Detection Based on Deep Learning and Belief Function

Cited by 6 publications

References 14 publications

Multimodal Co-Learning for Building Change Detection: A Domain Adaptation Framework Using VHR Images and Digital Surface Models

Multimodal Co-Learning for Building Change Detection: A Domain Adaptation Framework Using VHR Images and Digital Surface Models

Building Change Detection Based on 3D Co-Segmentation Using Satellite Stereo Imagery

Rapid identification of damaged buildings using incremental learning with transferred data from historical natural disaster cases

Contact Info

Product

Resources

About