scholarly journals ResUNet-a: A deep learning framework for semantic segmentation of remotely sensed data

2020 ◽  
Vol 162 ◽  
pp. 94-114 ◽  
Author(s):  
Foivos I. Diakogiannis ◽  
François Waldner ◽  
Peter Caccetta ◽  
Chen Wu
Keyword(s):  
Deep Learning ◽  
Semantic Segmentation ◽  
Remotely Sensed ◽  
Remotely Sensed Data ◽  
Learning Framework

Deep learning for remotely sensed data

2018 ◽  
Vol 145 ◽  
pp. 1-2 ◽  
Author(s):  
Giorgos Mountrakis ◽  
Jun Li ◽  
Xiaoqiang Lu ◽  
Olaf Hellwich
Keyword(s):  
Deep Learning ◽  
Remotely Sensed ◽  
Remotely Sensed Data

scholarly journals The Use of Remotely Sensed Data and Polish NFI Plots for Prediction of Growing Stock Volume Using Different Predictive Methods

2020 ◽  
Vol 12 (20) ◽  
pp. 3331
Author(s):  
Paweł Hawryło ◽  
Saverio Francini ◽  
Gherardo Chirici ◽  
Francesca Giannetti ◽  
Karolina Parkitna ◽  
...  
Keyword(s):  
Deep Learning ◽  
Remotely Sensed ◽  
Parameter Estimates ◽  
Remotely Sensed Data ◽  
Positional Accuracy ◽  
Growing Stock ◽  
3D Point Clouds ◽  
Predictive Methods ◽  
Predictive Approaches ◽  
Stock Volume

Forest growing stock volume (GSV) is an important parameter in the context of forest resource management. National Forest Inventories (NFIs) are routinely used to estimate forest parameters, including GSV, for national or international reporting. Remotely sensed data are increasingly used as a source of auxiliary information for NFI data to improve the spatial precision of forest parameter estimates. In this study, we combine data from the NFI in Poland with satellite images of Landsat 7 and 3D point clouds collected with airborne laser scanning (ALS) technology to develop predictive models of GSV. We applied an area-based approach using 13,323 sample plots measured within the second cycle of the NFI in Poland (2010–2014) with poor positional accuracy from several to 15 m. Four different predictive approaches were evaluated: multiple linear regression, k-Nearest Neighbours, Random Forest and Deep Learning fully connected neural network. For each of these predictive methods, three sets of predictors were tested: ALS-derived, Landsat-derived and a combination of both. The developed models were validated at the stand level using field measurements from 360 reference forest stands. The best accuracy (RMSE% = 24.2%) and lowest systematic error (bias% = −2.2%) were obtained with a deep learning approach when both ALS- and Landsat-derived predictors were used. However, the differences between the evaluated predictive approaches were marginal when using the same set of predictor variables. Only a slight increase in model performance was observed when adding the Landsat-derived predictors to the ALS-derived ones. The obtained results showed that GSV can be predicted at the stand level with relatively low bias and reasonable accuracy for coniferous species, even using field sample plots with poor positional accuracy for model development. Our findings are especially important in the context of GSV prediction in areas where NFI data are available but the collection of accurate positions of field plots is not possible or justified because of economic reasons.

scholarly journals A Novel Object-Based Deep Learning Framework for Semantic Segmentation of Very High-Resolution Remote Sensing Data: Comparison with Convolutional and Fully Convolutional Networks

2019 ◽  
Vol 11 (6) ◽  
pp. 684 ◽  
Author(s):  
Maria Papadomanolaki ◽  
Maria Vakalopoulou ◽  
Konstantinos Karantzalos
Keyword(s):  
Deep Learning ◽  
State Of The Art ◽  
Semantic Segmentation ◽  
Novel Object ◽  
Convolutional Networks ◽  
Learning Framework ◽  
Fully Convolutional Networks ◽  
Object Based ◽  
Deep Networks ◽  
Very High

Deep learning architectures have received much attention in recent years demonstrating state-of-the-art performance in several segmentation, classification and other computer vision tasks. Most of these deep networks are based on either convolutional or fully convolutional architectures. In this paper, we propose a novel object-based deep-learning framework for semantic segmentation in very high-resolution satellite data. In particular, we exploit object-based priors integrated into a fully convolutional neural network by incorporating an anisotropic diffusion data preprocessing step and an additional loss term during the training process. Under this constrained framework, the goal is to enforce pixels that belong to the same object to be classified at the same semantic category. We compared thoroughly the novel object-based framework with the currently dominating convolutional and fully convolutional deep networks. In particular, numerous experiments were conducted on the publicly available ISPRS WGII/4 benchmark datasets, namely Vaihingen and Potsdam, for validation and inter-comparison based on a variety of metrics. Quantitatively, experimental results indicate that, overall, the proposed object-based framework slightly outperformed the current state-of-the-art fully convolutional networks by more than 1% in terms of overall accuracy, while intersection over union results are improved for all semantic categories. Qualitatively, man-made classes with more strict geometry such as buildings were the ones that benefit most from our method, especially along object boundaries, highlighting the great potential of the developed approach.

scholarly journals BUILDING OUTLINE EXTRACTION FROM AERIAL IMAGERY AND DIGITAL SURFACE MODEL WITH A FRAME FIELD LEARNING FRAMEWORK

2021 ◽  
Vol XLIII-B2-2021 ◽  
pp. 487-493
Author(s):  
X. Sun ◽  
W. Zhao ◽  
R. V. Maretto ◽  
C. Persello
Keyword(s):  
Deep Learning ◽  
Semantic Segmentation ◽  
Aerial Images ◽  
Surface Model ◽  
Digital Surface Model ◽  
Frame Field ◽  
Learning Framework ◽  
Elevation Data ◽  
3D Information ◽  
Direction Information

Abstract. Deep learning-based semantic segmentation models for building delineation face the challenge of producing precise and regular building outlines. Recently, a building delineation method based on frame field learning was proposed by Girard et al. (2020) to extract regular building footprints as vector polygons directly from aerial RGB images. A fully convolution network (FCN) is trained to learn simultaneously the building mask, contours, and frame field followed by a polygonization method. With the direction information of the building contours stored in the frame field, the polygonization algorithm produces regular outlines accurately detecting edges and corners. This paper investigated the contribution of elevation data from the normalized digital surface model (nDSM) to extract accurate and regular building polygons. The 3D information provided by the nDSM overcomes the aerial images’ limitations and contributes to distinguishing the buildings from the background more accurately. Experiments conducted in Enschede, the Netherlands, demonstrate that the nDSM improves building outlines’ accuracy, resulting in better-aligned building polygons and prevents false positives. The investigated deep learning approach (fusing RGB + nDSM) results in a mean intersection over union (IOU) of 0.70 in the urban area. The baseline method (using RGB only) results in an IOU of 0.58 in the same area. A qualitative analysis of the results shows that the investigated model predicts more precise and regular polygons for large and complex structures.

scholarly journals Automatic Detection of Impervious Surfaces from Remotely Sensed Data Using Deep Learning

2021 ◽  
Vol 13 (16) ◽  
pp. 3166
Author(s):  
Jash R. Parekh ◽  
Ate Poortinga ◽  
Biplov Bhandari ◽  
Timothy Mayer ◽  
David Saah ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Large Scale ◽  
Mean Squared Error ◽  
Remotely Sensed ◽  
Landsat 8 ◽  
Impervious Surfaces ◽  
Remotely Sensed Data ◽  
Statistical Indices ◽  
Deep Learning Neural Network

The large scale quantification of impervious surfaces provides valuable information for urban planning and socioeconomic development. Remote sensing and GIS techniques provide spatial and temporal information of land surfaces and are widely used for modeling impervious surfaces. Traditionally, these surfaces are predicted by computing statistical indices derived from different bands available in remotely sensed data, such as the Landsat and Sentinel series. More recently, researchers have explored classification and regression techniques to model impervious surfaces. However, these modeling efforts are limited due to lack of labeled data for training and evaluation. This in turn requires significant effort for manual labeling of data and visual interpretation of results. In this paper, we train deep learning neural networks using TensorFlow to predict impervious surfaces from Landsat 8 images. We used OpenStreetMap (OSM), a crowd-sourced map of the world with manually interpreted impervious surfaces such as roads and buildings, to programmatically generate large amounts of training and evaluation data, thus overcoming the need for manual labeling. We conducted extensive experimentation to compare the performance of different deep learning neural network architectures, optimization methods, and the set of features used to train the networks. The four model configurations labeled U-Net_SGD_Bands, U-Net_Adam_Bands, U-Net_Adam_Bands+SI, and VGG-19_Adam_Bands+SI resulted in a root mean squared error (RMSE) of 0.1582, 0.1358, 0.1375, and 0.1582 and an accuracy of 90.87%, 92.28%, 92.46%, and 90.11%, respectively, on the test set. The U-Net_Adam_Bands+SI Model, similar to the others mentioned above, is a deep learning neural network that combines Landsat 8 bands with statistical indices. This model performs the best among all four on statistical accuracy and produces qualitatively sharper and brighter predictions of impervious surfaces as compared to the other models.

Sign in / Sign up

Export Citation Format

Share Document