scholarly journals Identifying Soil Erosion Processes in Alpine Grasslands on Aerial Imagery with a U-Net Convolutional Neural Network

2020 ◽  
Vol 12 (24) ◽  
pp. 4149
Author(s):  
Maxim Samarin ◽  
Lauren Zweifel ◽  
Volker Roth ◽  
Christine Alewell
Keyword(s):  
Neural Network ◽  
Land Use ◽  
Soil Erosion ◽  
Convolutional Neural Network ◽  
Temporal Trends ◽  
Aerial Imagery ◽  
Aerial Images ◽  
Quality Data ◽  
Alpine Grasslands ◽  
Erosion Processes

Erosion in alpine grasslands is a major threat to ecosystem services of alpine soils. Natural causes for the occurrence of soil erosion are steep topography and prevailing climate conditions in combination with soil fragility. To increase our understanding of ongoing erosion processes and support sustainable land-use management, there is a need to acquire detailed information on spatial occurrence and temporal trends. Existing approaches to identify these trends are typically laborious, have lack of transferability to other regions, and are consequently only applicable to smaller regions. In order to overcome these limitations and create a sophisticated erosion monitoring tool capable of large-scale analysis, we developed a model based on U-Net, a fully convolutional neural network, to map different erosion processes on high-resolution aerial images (RGB, 0.25–0.5 m). U-Net was trained on a high-quality data set consisting of labeled erosion sites mapped with object-based image analysis (OBIA) for the Urseren Valley (Central Swiss Alps) for five aerial images (16 year period). We used the U-Net model to map the same study area and conduct quality assessments based on a held-out test region and a temporal transferability test on new images. Erosion classes are assigned according to their type (shallow landslide and sites with reduced vegetation affected by sheet erosion) or land-use impacts (livestock trails and larger management affected areas). We show that results obtained by OBIA and U-Net follow similar linear trends for the 16 year study period, exhibiting increases in total degraded area of 167% and 201%, respectively. Segmentations of eroded sites are generally in good agreement, but also display method-specific differences, which lead to an overall precision of 73%, a recall of 84%, and a F1-score of 78%. Our results show that U-Net is transferable to spatially (within our study area) and temporally unseen data (data from new years) and is therefore a method suitable to efficiently and successfully capture the temporal trends and spatial heterogeneity of degradation in alpine grasslands. Additionally, U-Net is a powerful and robust tool to map erosion sites in a predictive manner utilising large amounts of new aerial imagery.

Identification of Soil Erosion in Alpine Grasslands on High-Resolution Aerial Images: Switching from Object-based Image Analysis to Deep Learning?

2020 ◽  
Author(s):  
Lauren Zweifel ◽  
Maxim Samarin ◽  
Katrin Meusburger ◽  
Volker Roth ◽  
Christine Alewell
Keyword(s):  
Land Use ◽  
Image Analysis ◽  
Soil Erosion ◽  
Soil Degradation ◽  
Aerial Images ◽  
Alpine Grasslands ◽  
Object Based Image Analysis ◽  
Erosion Processes ◽  
Object Based ◽  
Land Use Practices

<p>Soil erosion in Alpine grassland areas is an ecological threat caused by the extreme topography, prevailing climate conditions and land-use practices but enhanced by climate change (e.g., heavy precipitation events, changing snow dynamics) in combination with changing land-use practices (e.g, more intensely used pastures). To increase our understanding of ongoing soil erosion processes in Alpine grasslands, there is a need to acquire detailed information on spatial extension and temporal trends.</p><p>In the past, we have successfully applied a semi-automatic method using an object-based image analysis (OBIA) framework with high-resolution aerial images (0.25-0.5m) and a digital terrain model (2m) to map erosion features in the Central Swiss Alps (Urseren Valley, Canton Uri, Switzerland). Degraded sites are classified according to the major erosion process (shallow landslides; sites with reduced vegetation cover affected by sheet erosion) or triggering factors (trampling by livestock; management effects) (Zweifel et al. 2019). We now aim to apply a deep learning (DL) model with the purpose of fast and efficient spatial upscaling(e.g., alpine-wide analysis). While OBIA yields high quality results, there are multiple constraints, such as labor-intensive steps and the requirement of expert knowledge, which make the method unsuitable for larger scale applications. The results of OBIA are used as a training dataset for our DL model. The DL approach uses fully-convolutional networks with the U-Net architecture and is capable of rapid segmentation and classification to identify areas with reduced vegetation cover and bare soil sites.</p><p>Results for the Urseren Valley (Canton Uri, Switzerland) show an increase in total area affected by soil degradation of 156 ±18% during a 16-year observation period (2000-2016). A comparison of the two methods (OBIA and DL) shows that DL results for the Urseren Valley follow similar trends for the 16-year period and that the segmentations of eroded sites are in good agreement (IoU = 0.83). First transferability tests to other valleys not considered during training of the DL model are very promising, confirming that DL is a well-suited and efficient method for future projects to map and assess soil erosion processes in grassland areas at regional scales.</p><p> </p><p><strong>References</strong></p><p>L. Zweifel, K. Meusburger, and C. Alewell. Spatio-temporal pattern of soil degradation in a Swiss Alpine grassland catchment. Remote Sensing of Environment, 235, 2019.</p>

scholarly journals Multiple Pedestrians and Vehicles Tracking in Aerial Imagery Using a Convolutional Neural Network

2021 ◽  
Vol 13 (10) ◽  
pp. 1953
Author(s):  
Seyed Majid Azimi ◽  
Maximilian Kraus ◽  
Reza Bahmanyar ◽  
Peter Reinartz
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Object Tracking ◽  
Short Term Memory ◽  
Aerial Imagery ◽  
Future Research ◽  
Short Term ◽  
Term Memory ◽  
Long Short Term Memory

In this paper, we address various challenges in multi-pedestrian and vehicle tracking in high-resolution aerial imagery by intensive evaluation of a number of traditional and Deep Learning based Single- and Multi-Object Tracking methods. We also describe our proposed Deep Learning based Multi-Object Tracking method AerialMPTNet that fuses appearance, temporal, and graphical information using a Siamese Neural Network, a Long Short-Term Memory, and a Graph Convolutional Neural Network module for more accurate and stable tracking. Moreover, we investigate the influence of the Squeeze-and-Excitation layers and Online Hard Example Mining on the performance of AerialMPTNet. To the best of our knowledge, we are the first to use these two for regression-based Multi-Object Tracking. Additionally, we studied and compared the L1 and Huber loss functions. In our experiments, we extensively evaluate AerialMPTNet on three aerial Multi-Object Tracking datasets, namely AerialMPT and KIT AIS pedestrian and vehicle datasets. Qualitative and quantitative results show that AerialMPTNet outperforms all previous methods for the pedestrian datasets and achieves competitive results for the vehicle dataset. In addition, Long Short-Term Memory and Graph Convolutional Neural Network modules enhance the tracking performance. Moreover, using Squeeze-and-Excitation and Online Hard Example Mining significantly helps for some cases while degrades the results for other cases. In addition, according to the results, L1 yields better results with respect to Huber loss for most of the scenarios. The presented results provide a deep insight into challenges and opportunities of the aerial Multi-Object Tracking domain, paving the way for future research.

scholarly journals Feature Importance Analysis for Local Climate Zone Classification Using a Residual Convolutional Neural Network with Multi-Source Datasets

2018 ◽  
Vol 10 (10) ◽  
pp. 1572 ◽  
Author(s):  
Chunping Qiu ◽  
Michael Schmitt ◽  
Lichao Mou ◽  
Pedram Ghamisi ◽  
Xiao Zhu
Keyword(s):  
Neural Network ◽  
Land Use ◽  
Convolutional Neural Network ◽  
Large Scale ◽  
Landsat 8 ◽  
Climate Zone ◽  
Local Climate ◽  
Feature Importance ◽  
Local Climate Zone ◽  
Sentinel 2

Global Local Climate Zone (LCZ) maps, indicating urban structures and land use, are crucial for Urban Heat Island (UHI) studies and also as starting points to better understand the spatio-temporal dynamics of cities worldwide. However, reliable LCZ maps are not available on a global scale, hindering scientific progress across a range of disciplines that study the functionality of sustainable cities. As a first step towards large-scale LCZ mapping, this paper tries to provide guidance about data/feature choice. To this end, we evaluate the spectral reflectance and spectral indices of the globally available Sentinel-2 and Landsat-8 imagery, as well as the Global Urban Footprint (GUF) dataset, the OpenStreetMap layers buildings and land use and the Visible Infrared Imager Radiometer Suite (VIIRS)-based Nighttime Light (NTL) data, regarding their relevance for discriminating different Local Climate Zones (LCZs). Using a Residual convolutional neural Network (ResNet), a systematic analysis of feature importance is performed with a manually-labeled dataset containing nine cities located in Europe. Based on the investigation of the data and feature choice, we propose a framework to fully exploit the available datasets. The results show that GUF, OSM and NTL can contribute to the classification accuracy of some LCZs with relatively few samples, and it is suggested that Landsat-8 and Sentinel-2 spectral reflectances should be jointly used, for example in a majority voting manner, as proven by the improvement from the proposed framework, for large-scale LCZ mapping.

scholarly journals Consideration of land use in the evaluation of erosive processes in gully erosion

2020 ◽  
Vol 24 ◽  
pp. e25
Author(s):  
Tamara Vieira Pascoto ◽  
Simone Andrea Furegatti ◽  
Anna Silvia Palcheco Peixoto
Keyword(s):  
Land Use ◽  
Visual Analysis ◽  
Aerial Images ◽  
Aerial Photos ◽  
City Hall ◽  
Erosion Processes ◽  
Main Factors ◽  
The City ◽  
Erosive Feature

There are several factors that directly or indirectly influence erosion processes. In order to properly understand the behavior of these processes, some factors need to be analyzed together. Determining them wrongly can compromise the study resulting in wrong actions. For this reason, methodologies are always sought to measure them quantitatively and qualitatively in the most accurate possible way. Land use is one of the main factors liable to inaccuracies in its determination. To use this parameter in mapping erosive processes, researchers need to delimit it, classify it, and measure it. In order to better understand the complexity of considering this parameter, the present study analyzed an erosive feature that, although stabilized, has a component in constant development. Initially, a visual analysis indicated the same classification of land use for both conditions, despite having different behaviors, leading to the need for a detailed analysis. Such analysis comprised a historical survey through aerial photos and interviews with residents and employees of the city hall about the evolution of the feature from 2008 to 2019. It also included the analysis of other influencing factors that could be responsible for this difference in behavior in the area. Two different traces of the contribution areas of the gully and branch were also considered. One considering only aerial images, and the other considering the knowledge acquired during the research about the evolution of the feature. It was concluded, then, that an analysis of the use-only occupation factor based on aerial images can accentuate the inaccuracy of the measurement of this factor.

scholarly journals Deep Learning-Based Identification of Collapsed, Non-Collapsed and Blue Tarp-Covered Buildings from Post-Disaster Aerial Images

2020 ◽  
Vol 12 (12) ◽  
pp. 1924 ◽  
Author(s):  
Hiroyuki Miura ◽  
Tomohiro Aridome ◽  
Masashi Matsuoka
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Convolutional Neural Network ◽  
Building Damage ◽  
Aerial Images ◽  
Automated Identification ◽  
Damage Data ◽  
Collapsed Buildings ◽  
Post Disaster ◽  
Earthquake Data

A methodology for the automated identification of building damage from post-disaster aerial images was developed based on convolutional neural network (CNN) and building damage inventories. The aerial images and the building damage data obtained in the 2016 Kumamoto, and the 1995 Kobe, Japan earthquakes were analyzed. Since the roofs of many moderately damaged houses are covered with blue tarps immediately after disasters, not only collapsed and non-collapsed buildings but also the buildings covered with blue tarps were identified by the proposed method. The CNN architecture developed in this study correctly classifies the building damage with the accuracy of approximately 95 % in both earthquake data. We applied the developed CNN model to aerial images in Chiba, Japan, damaged by the typhoon in September 2019. The result shows that more than 90 % of the building damage are correctly classified by the CNN model.

scholarly journals Estimating Human Impacts on Soil Erosion Considering Different Hillslope Inclinations and Land Uses in the Coastal Region of Syria

Water ◽  
2020 ◽  
Vol 12 (10) ◽  
pp. 2786 ◽  
Author(s):  
Safwan Mohammed ◽  
Hazem G. Abdo ◽  
Szilard Szabo ◽  
Quoc Bao Pham ◽  
Imre J. Holb ◽  
...  
Keyword(s):  
Land Use ◽  
Soil Erosion ◽  
Land Cover ◽  
Soil Loss ◽  
Agricultural Land ◽  
Human Impacts ◽  
Coastal Region ◽  
Water Erosion ◽  
Land Uses ◽  
Erosion Processes

Soils in the coastal region of Syria (CRoS) are one of the most fragile components of natural ecosystems. However, they are adversely affected by water erosion processes after extreme land cover modifications such as wildfires or intensive agricultural activities. The main goal of this research was to clarify the dynamic interaction between erosion processes and different ecosystem components (inclination, land cover/land use, and rainy storms) along with the vulnerable territory of the CRoS. Experiments were carried out in five different locations using a total of 15 erosion plots. Soil loss and runoff were quantified in each experimental plot, considering different inclinations and land uses (agricultural land (AG), burnt forest (BF), forest/control plot (F)). Observed runoff and soil loss varied greatly according to both inclination and land cover after 750 mm of rainfall (26 events). In the cultivated areas, the average soil water erosion ranged between 0.14 ± 0.07 and 0.74 ± 0.33 kg/m2; in the BF plots, mean soil erosion ranged between 0.03 ± 0.01 and 0.24 ± 0.10 kg/m2. The lowest amount of erosion was recorded in the F plots where the erosion ranged between 0.1 ± 0.001 and 0.07 ± 0.03 kg/m2. Interestingly, the General Linear Model revealed that all factors (i.e., inclination, rainfall and land use) had a significant (p < 0.001) effect on the soil loss. We concluded that human activities greatly influenced soil erosion rates, being higher in the AG lands, followed by BF and F. Therefore, the current study could be very useful to policymakers and planners for proposing immediate conservation or restoration plans in a less studied area which has been shown to be vulnerable to soil erosion processes.

Sign in / Sign up

Export Citation Format

Share Document