Arbitrary-Shaped Building Boundary-Aware Detection With Pixel Aggregation Network

Jiang, Xin; Zhang, Xinchang; Xin, Qinchuan; Xu, Xi; Zhang, Pengcheng

doi:10.1109/jstars.2020.3017934

Cited by 13 publications

(9 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…1(b). Jiang et al [14] also pointed out that in the task of building extraction, different categories of boundary recognition capabilities are different, which will have a certain impact on the final segmentation accuracy.…”

Section: Introductionmentioning

confidence: 99%

“…Second, used morphological filtering to optimize building boundaries, and improve boundary regularity. Similar to our method, Jiang et al [14] proposed a predictive optimization architecture, which consists of an encoder-decoder network and residual refinement modules responsible for prediction and refinement. To enhance the buildings expression ability, the authors also introduces a composite loss function.…”

Section: Introductionmentioning

confidence: 99%

“…Then, the D-Net is used to refine the results of E-Net and get a prediction with higher detail quality. Different from [14] using different loss functions to make the network pay attention to the significance map of the different-level pixels, we propose a new fusion strategy to ensembling the outputs of the two networks in a weighted manner. Experimental results on Inria Aerial Image Labeling dataset and the ISPRS Vaihingen 2D semantic labeling dataset demonstrate that the performance of E-D-Net is competitive with that of several state-of-the-art methods.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

E-D-Net: Automatic Building Extraction From High-Resolution Aerial Images With Boundary Information

Zhu

Liang

Yan

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

The automatic extraction of buildings from highresolution aerial imagery plays a significant role in many urban applications. Recently, convolution neural network (CNN) has gained much attention in remote sensing field and achieved remarkable performance in building segmentation from visible aerial images. However, most of the existing CNN-based methods still have the problem of tending to produce predictions with poor boundaries. To address this problem, in this paper a novel semantic segmentation neural network named Edge-Detail-Network (E-D-Net) is proposed for building segmentation from visible aerial images. The proposed E-D-Net consists of two subnetworks E-Net and D-Net. On the one hand, E-Net is designed to capture and preserve the edge information of the images. On the other hand, D-Net is designed to refine the results of E-Net and get a prediction with higher detail quality. Furthermore, a novel fusion strategy which combines the outputs of the two sub-networks is proposed to integrate edge information with fine details. Experimental results on the Inria Aerial Image Labeling Dataset and the ISPRS Vaihingen 2D semantic labeling dataset demonstrate that, compared with the existing CNN-based model, the proposed E-D-Net provides noticeably more robust and higher building extraction performance, thus making it a useful tool for practical application scenarios.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

E-D-Net: Automatic Building Extraction From High-Resolution Aerial Images With Boundary Information

Zhu

Liang

Yan

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Moreover, the fully convolutional networks (FCN) derived from CNNs are capable of making pixel-wise classifications [18]. However, due to the large intra-class variance and low inter-class variance of remote sensing images, it has been proven that directly applying deep methods designed for natural images to remote sensing images leads to accuracy drop [2], [19], [20].…”

Section: Introductionmentioning

confidence: 99%

Scale-Robust Deep-Supervision Network for Mapping Building Footprints From High-Resolution Remote Sensing Images

Guo

Tang

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Building footprint information is one of the key factors for sustainable urban planning and environmental monitoring. Mapping building footprints from remote sensing images is an important and challenging task in the earth observation field. Over the years, convolutional neural networks have shown outstanding improvements in the building extraction field due to their ability to automatically extract hierarchical features and make building predictions. However, as buildings are various in different sizes, scenes, and roofing materials, it is hard to precisely depict buildings of varied sizes, especially in large areas (e.g. nationwide). To tackle these limitations, we propose a novel deep-supervision convolutional neural network (denoted as DS-Net) for extracting building footprints from high-resolution remote sensing images. In the proposed network, we applied deep supervision with an extra lightweight encoder, which enables the network to learn representative building features of different scales. Furthermore, a scale attention module is designed to aggregate multiscale features and generate the final building prediction. Experiments on two publicly available building datasets, including the WHU Building Dataset and the Massachusetts Building Dataset, show the effectiveness of the proposed method. With only a 0.22M increment of parameters compared with U-Net, the proposed DS-Net achieved an IoU of 90.4% on the WHU Building Dataset and 73.8% on the Massachusetts Dataset. DS-Net also outperforms the state-of-theart building extraction methods on the two datasets, indicating the effectiveness of the proposed deep supervision and scale attention. Index Terms-Remote sensing image, deep learning, convolutional neural network, building footprint extraction.

show abstract

“…To detect dense and small buildings, Shu et al [34] proposed an endto-end model guided by center point. Jiang et al [35] used an encoder-decoder network and residual refinement module to form a predictive architecture for the prediction. Yao et al [36] applied the visual saliency and Condition Random Field (CRF) to train a coarse-to-fine model.…”

mentioning

confidence: 99%

Tilt Correction Toward Building Detection of Remote Sensing Images

Liu

Jiang

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Building detection is a crucial task in the field of remote sensing, and which can facilitate urban construction planning, disaster survey, and emergency landing. However, for large-size remote sensing images, the great majority of existing works have ignored the image tilt problem. This problem can result in partitioning buildings into separately oblique parts when the large-size images are partitioned. This is not beneficial to preserve semantic completeness of the building objects. Motivated by the fact above, we firstly propose a framework that detecting objects in large-size image, particularly for the building detection. The framework mainly consists of two phases. In the first phase, we particularly propose a Tilt Correction (TC) algorithm which contains three steps: texture mapping, tilt angle assessment, and image rotation. In the second phase, the building detection is performed with object detectors, especially deep neural network based methods. Last but not least, the detection results will be inversely mapped to the original large-size image. Furthermore, a challenging dataset named Aerial Image Building Detection (AIBD) is contributed for the public research. To evaluate the TC method, we also define an evaluation metric to compute the Cost of Building Partition (CoBP). The experiment results demonstrate the effects of the proposed method for the building detection.

show abstract

Arbitrary-Shaped Building Boundary-Aware Detection With Pixel Aggregation Network

Cited by 13 publications

References 55 publications

E-D-Net: Automatic Building Extraction From High-Resolution Aerial Images With Boundary Information

E-D-Net: Automatic Building Extraction From High-Resolution Aerial Images With Boundary Information

Scale-Robust Deep-Supervision Network for Mapping Building Footprints From High-Resolution Remote Sensing Images

Tilt Correction Toward Building Detection of Remote Sensing Images

Contact Info

Product

Resources

About