Water Body Extraction From Very High-Resolution Remote Sensing Imagery Using Deep U-Net and a Superpixel-Based Conditional Random Field Model

Feng, Wenqing; Sui, Haigang; Huang, Weiming; Xu, Chuan; An, Kaiqiang

doi:10.1109/lgrs.2018.2879492

Cited by 136 publications

(67 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…U-Net has achieved state-of-the-art results on the ISBI EM segmentation challenge [22] and the ISBI cell tracking challenge [23]. The U-Net has also been successfully employed in remote sensing applications, such as the detection of features [24], the extraction of water bodies from satellite imagery [25], and the extraction of buildings from aerial imagery [26].…”

Section: Network Architecturementioning

confidence: 99%

Automatic Mapping of Center Pivot Irrigation Systems from Satellite Images Using Deep Learning

Saraiva¹,

Protas²,

Salgado³

et al. 2020

Remote Sensing

View full text Add to dashboard Cite

The availability of freshwater is becoming a global concern. Because agricultural consumption has been increasing steadily, the mapping of irrigated areas is key for supporting the monitoring of land use and better management of available water resources. In this paper, we propose a method to automatically detect and map center pivot irrigation systems using U-Net, an image segmentation convolutional neural network architecture, applied to a constellation of PlanetScope images from the Cerrado biome of Brazil. Our objective is to provide a fast and accurate alternative to map center pivot irrigation systems with very high spatial and temporal resolution imagery. We implemented a modified U-Net architecture using the TensorFlow library and trained it on the Google cloud platform with a dataset built from more than 42,000 very high spatial resolution PlanetScope images acquired between August 2017 and November 2018. The U-Net implementation achieved a precision of 99% and a recall of 88% to detect and map center pivot irrigation systems in our study area. This method, proposed to detect and map center pivot irrigation systems, has the potential to be scaled to larger areas and improve the monitoring of freshwater use by agricultural activities.

show abstract

Section: Network Architecturementioning

confidence: 99%

Automatic Mapping of Center Pivot Irrigation Systems from Satellite Images Using Deep Learning

Saraiva¹,

Protas²,

Salgado³

et al. 2020

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Graph model-based methods are also an available way to obtain contextual information such as conditional random fields (CRFs). CRFs are generally used in postprocessing to refine spatial details [38]. Among those methods, CRFs only optimize the segmentation results and do not participate in the training process of the network.…”

Section: Semantic Segmentation For High-resolution Aerial Imagesmentioning

confidence: 99%

Urban Land Cover Classification of High-Resolution Aerial Imagery Using a Relation-Enhanced Multiscale Convolutional Network

Liu

Zeng

et al. 2020

Remote Sensing

View full text Add to dashboard Cite

Urban land cover classification for high-resolution images is a fundamental yet challenging task in remote sensing image analysis. Recently, deep learning techniques have achieved outstanding performance in high-resolution image classification, especially the methods based on deep convolutional neural networks (DCNNs). However, the traditional CNNs using convolution operations with local receptive fields are not sufficient to model global contextual relations between objects. In addition, multiscale objects and the relatively small sample size in remote sensing have also limited classification accuracy. In this paper, a relation-enhanced multiscale convolutional network (REMSNet) method is proposed to overcome these weaknesses. A dense connectivity pattern and parallel multi-kernel convolution are combined to build a lightweight and varied receptive field sizes model. Then, the spatial relation-enhanced block and the channel relation-enhanced block are introduced into the network. They can adaptively learn global contextual relations between any two positions or feature maps to enhance feature representations. Moreover, we design a parallel multi-kernel deconvolution module and spatial path to further aggregate different scales information. The proposed network is used for urban land cover classification against two datasets: the ISPRS 2D semantic labelling contest of Vaihingen and an area of Shanghai of about 143 km 2 . The results demonstrate that the proposed method can effectively capture long-range dependencies and improve the accuracy of land cover classification. Our model obtains an overall accuracy (OA) of 90.46% and a mean intersection-over-union (mIoU) of 0.8073 for Vaihingen and an OA of 88.55% and a mIoU of 0.7394 for Shanghai.Automatic urban land cover classification in remote sensing images is a difficult task due to several challenges. First of all, high-resolution remote sensing images with abundant details generally have characteristics of high intraclass variance and low inter-class variance [10]. For example, different classes of ground objects may have a similar appearance in remote sensing images, such as trees and low vegetation or roofs and roads. Meanwhile, there is occlusion between different objects. To solve this issue, context information has been widely studied in remote sensing image classification [11]. Contextual information refers to the dependencies between objects, such as cars on roads. Misclassification will be reduced when taking these context relationships into account. In addition, ground objects often have various scales in remote sensing images. Cars and buildings have large differences in size. It is difficult to simultaneously distinguish objects with distinct scales. Generally, deeper layers with larger receptive fields are more suited for segmenting large objects, while shallower layers with smaller receptive fields are suitable to segment small objects [12]. Therefore, it is necessary to enhance global context relationships and fuse multiscale information for pixel-...

show abstract

“…However, it is necessary to choose an initial point of the contour of the water, and automatic acquisition of this initial position is not frequently easily achieved. Furthermore, these manual decisions impact on the accuracy and efficiency of the detection of water bodies (Feng et al, 2019), (Hemalatha et al, 2018).…”

Section: State Of the Artmentioning

confidence: 99%

“…Although there exist some approaches using big data analysis and image processing algorithms to cope with these tasks (Ayma et al, 2016), most of the solutions still rely on the expertise of researchers to select the thresholds and determine which features are the most representative. For these reasons, in the last years, there are several investigations using convolution neural networks such as those reported in (Miao et al, 2018), (Nowaczynski, 2017), (Feng et al, 2019), (Talal et al, 2018), and (Hu et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Application of Semantic Segmentation With Few Labels in the Detection of Water Bodies From Perusat-1 Satellite’s Images

Gonzalez

Sankaran

Ayma

et al. 2020

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Abstract. Remote sensing is widely used to monitor earth surfaces with the main objective of extracting information from it. Such is the case of water surface, which is one of the most affected extensions when flood events occur, and its monitoring helps in the analysis of detecting such affected areas, considering that adequately defining water surfaces is one of the biggest problems that Peruvian authorities are concerned with. In this regard, semiautomatic mapping methods improve this monitoring, but this process remains a time-consuming task and into the subjectivity of the experts.In this work, we present a new approach for segmenting water surfaces from satellite images based on the application of convolutional neural networks. First, we explore the application of a U-Net model and then a transfer knowledge-based model. Our results show that both approaches are comparable when trained using an 680-labelled satellite image dataset; however, as the number of training samples is reduced, the performance of the transfer knowledge-based model, which combines high and very high image resolution characteristics, is improved.

show abstract

Water Body Extraction From Very High-Resolution Remote Sensing Imagery Using Deep U-Net and a Superpixel-Based Conditional Random Field Model

Cited by 136 publications

References 13 publications

Automatic Mapping of Center Pivot Irrigation Systems from Satellite Images Using Deep Learning

Automatic Mapping of Center Pivot Irrigation Systems from Satellite Images Using Deep Learning

Urban Land Cover Classification of High-Resolution Aerial Imagery Using a Relation-Enhanced Multiscale Convolutional Network

Application of Semantic Segmentation With Few Labels in the Detection of Water Bodies From Perusat-1 Satellite’s Images

Contact Info

Product

Resources

About