A Large-Scale Mapping Scheme for Urban Building From Gaofen-2 Images Using Deep Learning and Hierarchical Approach

Wang, Guizhou; He, Guojin; Yin, Ranyu; Long, Tao; Zhang, Zhaoming; Chen, Si‐Bao; Luo, Bin

doi:10.1109/jstars.2021.3123398

Cited by 13 publications

(5 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Even though this study comes with its limitations for not training with a much bigger dataset or much more complex deep-learning architecture like transformers [56] due to hardware and resource limitations it shows the capability of Attention-U-net in generating fractional built-up cover at a large scale. Attention-U-net exhibited great promise, especially in scenarios with limited built-up pixel coverage largely due to its attention block mechanism helping it distinguish the background (nonbuilt-up pixels) from the foreground (or, built-up pixels).…”

Section: Discussionmentioning

confidence: 98%

Deep Learning for Built-up Fractional Mapping using Sentinel-2 images: A Case Study in Delhi, India

Rawat,

Gupta,

Persello

2024

Preprint

View full text Add to dashboard Cite

In our increasingly urbanized world, precise and up-to-date maps of human settlements are essential for sustainable urban development policies. The availability of open-access Sentinel-2 data from the Copernicus program presents an opportunity to create a comprehensive global map of human settlements, offering a detailed view of built areas on a large scale. This study estimates large-scale built-up fractions using encoder-decoder deep learning architectures like U-net, Res-U-net, and Attention-U-net in the large and complex urban area of Delhi, India. Openly available datasets like Open Street Map (OSM) and Microsoft building footprint datasets are used to derive built-up fractions at 10×10m resolution cells for over 34,000 km2. Our results show that Attention-U-net with the Huber loss function performs the best in different built-up densities (i.e., urban, semi-urban or rural) with an R2 score of 0.631, while Res-U-net and U-net obtained an R2 score of 0.623 and 0.612, respectively. The investigated networks significantly improve the accuracy over the latest Global Human Settlement Layer product (GHSL-S2), which uses a deep-CNN and reaches an R2 of 0.387 in our case study area. The result of this study yields a valuable spatial layer for examining the spatial distribution of human settlements across the entire spectrum from rural to urban areas.

show abstract

Section: Discussionmentioning

confidence: 98%

Deep Learning for Built-up Fractional Mapping using Sentinel-2 images: A Case Study in Delhi, India

Rawat,

Gupta,

Persello

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Landsat is a collection of continuously obtained space‐based moderate‐resolution land remote sensing data; these datasets are available for download for free. The Gaofen‐2 is one of many satellites that China has launched to assist with high‐accuracy geographical mapping, land and resource surveying, environmental change monitoring, and real‐time observation for disaster prevention and mitigation, among other activities (Wang & He, 2021). There are additional websites where we can find single or numerous images associated with a specific date and time.…”

Section: Overview Of Satellite Imagesmentioning

confidence: 99%

“…The standard steps followed by the literature are depicted in Figure 9. Zhou et al (Wang & He, 2021) proposed a hierarchical approach for urban building mapping for change detection. Dense low‐rise buildings (DLB) and sparse independent buildings (SIB) are two types of urban structures.…”

Section: Urban Flood Prediction Approachesmentioning

confidence: 99%

“…Multispectral remote sensing entails the capture of images in multiple broad wavelength bands, including visible, near-infrared, and shortwave infrared (Kim & Duc, 2020;Kalinicheva et al, 2019;Jacobson, 2011;Nguyen & Bae, 2020;Peng et al, 2019Peng et al, , 2021Song et al, 2021;Wang & He, 2021). A multi-spectral image is a collection of many monochrome images captured with different sensors of the same scene.…”

Section: Multispectral Imagesmentioning

confidence: 99%

See 1 more Smart Citation

A review of intelligent models for mapping city development and urban flooding

Senthilkumar

Arthur

2023

Land Degrad Dev

View full text Add to dashboard Cite

With urbanization is on the rise, it wreaks havoc on the hydrological processes within a catchment, resulting in a deteriorating water environment. Floods can greatly raise the cost of mitigation efforts due to their damaging consequences. The natural water balance is disrupted by an increase in impermeable areas. As runoff rises, so do flood peaks and volumes even in short, low‐intensity showers. Urban flooding is a significant impact of urbanization that has drawn the attention of experts. Currently, there are techniques based on distributed hydrologic modeling and Geographic Information Systems (GIS) used to assess the influence of land use for urbanization on the environment. Deep learning techniques with their immense performance in Computer Vision and Image Processing fields have been extended to the problem of urban floods by integrating them with a GIS‐based methodology. This study outlines multiple techniques that have been used to address urban floods.

show abstract

“…Additionally, some pre-trained models and open-source building datasets, (e.g., WHU building dataset I and II, WHU aerial dataset, Aerial imagery for roof segmentation, SpaceNet 1 and 2, Inria Aerial Image Labeling Dataset, and Massachusetts building dataset) could be used to train CNN models for building extraction (Table 1). Despite this, it still requires a large amount of labeled data to calculate and update the parameters of CNN layers for achieving optimal performance in a particular area or large-scale geographic region due to differences in landscapes, building diversity, and the elements associated with sensors, acquisition times and spatial resolutions [12][13][14]. [17,19,20] Inria Aerial Image Labeling Dataset 0.3 360 5000 × 5000 10 cities worldwide [21] Massachusetts Buildings Dataset 1 151 1500 × 1500 Massachusetts, USA [22] In real-world applications, developing an appropriate CNN model and generating a building map for a specific study area mainly includes four stages: (1) data acquisition and preprocessing, (2) data annotation, (3) model training and evaluation, (4) generation of building map [23][24][25].…”

Section: Introductionmentioning

confidence: 99%

Suggestive Data Annotation for CNN-Based Building Footprint Mapping Based on Deep Active Learning and Landscape Metrics

Zhang

Dong

2022

Remote Sensing

View full text Add to dashboard Cite

Convolutional neural network (CNN)-based very high-resolution (VHR) image segmentation has become a common way of extracting building footprints. Despite publicly available building datasets and pre-trained CNN models, it is still necessary to prepare sufficient labeled image tiles to train CNN models from scratch or update the parameters of pre-trained CNN models to extract buildings accurately in real-world applications, especially the large-scale building extraction, due to differences in landscapes and data sources. Deep active learning is an effective technique for resolving this issue. This study proposes a framework integrating two state-of-the-art (SOTA) models, U-Net and DeeplabV3+, three commonly used active learning strategies, (i.e., margin sampling, entropy, and vote entropy), and landscape characterization to illustrate the performance of active learning in reducing the effort of data annotation, and then understand what kind of image tiles are more advantageous for CNN-based building extraction. The framework enables iteratively selecting the most informative image tiles from the unlabeled dataset for data annotation, training the CNN models, and analyzing the changes in model performance. It also helps us to understand the landscape features of iteratively selected image tiles via active learning by considering building as the focal class and computing the percent, the number of patches, edge density, and landscape shape index of buildings based on labeled tiles in each selection. The proposed method was evaluated on two benchmark building datasets, WHU satellite dataset II and WHU aerial dataset. Models in each iteration were trained from scratch on all labeled tiles. Experimental results based on the two datasets indicate that, for both U-Net and DeeplabV3+, the three active learning strategies can reduce the number of image tiles to be annotated and achieve good model performance with fewer labeled image tiles. Moreover, image tiles with more building patches, larger areas of buildings, longer edges of buildings, and more dispersed building distribution patterns were more effective for model training. The study not only provides a framework to reduce the data annotation efforts in CNN-based building extraction but also summarizes the preliminary suggestions for data annotation, which could facilitate and guide data annotators in real-world applications.

show abstract

A Large-Scale Mapping Scheme for Urban Building From Gaofen-2 Images Using Deep Learning and Hierarchical Approach

Cited by 13 publications

References 57 publications

Deep Learning for Built-up Fractional Mapping using Sentinel-2 images: A Case Study in Delhi, India

Deep Learning for Built-up Fractional Mapping using Sentinel-2 images: A Case Study in Delhi, India

A review of intelligent models for mapping city development and urban flooding

Suggestive Data Annotation for CNN-Based Building Footprint Mapping Based on Deep Active Learning and Landscape Metrics

Contact Info

Product

Resources

About