Mapping deforestation is an essential step in the process of managing tropical rainforests. It lets us understand and monitor both legal and illegal deforestation and its implications, which include ...the effect deforestation may have on climate change through greenhouse gas emissions. Given that there is ample room for improvements when it comes to mapping deforestation using satellite imagery, in this study, we aimed to test and evaluate the use of algorithms belonging to the growing field of deep learning (DL), particularly convolutional neural networks (CNNs), to this end. Although studies have been using DL algorithms for a variety of remote sensing tasks for the past few years, they are still relatively unexplored for deforestation mapping. We attempted to map the deforestation between images approximately one year apart, specifically between 2017 and 2018 and between 2018 and 2019. Three CNN architectures that are available in the literature—SharpMask, U-Net, and ResUnet—were used to classify the change between years and were then compared to two classic machine learning (ML) algorithms—random forest (RF) and multilayer perceptron (MLP)—as points of reference. After validation, we found that the DL models were better in most performance metrics including the Kappa index, F1 score, and mean intersection over union (mIoU) measure, while the ResUnet model achieved the best overall results with a value of 0.94 in all three measures in both time sequences. Visually, the DL models also provided classifications with better defined deforestation patches and did not need any sort of post-processing to remove noise, unlike the ML models, which needed some noise removal to improve results.
The Synthetic Aperture Radar (SAR) time series allows describing the rice phenological cycle by the backscattering time signature. Therefore, the advent of the Copernicus Sentinel-1 program expands ...studies of radar data (C-band) for rice monitoring at regional scales, due to the high temporal resolution and free data distribution. Recurrent Neural Network (RNN) model has reached state-of-the-art in the pattern recognition of time-sequenced data, obtaining a significant advantage at crop classification on the remote sensing images. One of the most used approaches in the RNN model is the Long Short-Term Memory (LSTM) model and its improvements, such as Bidirectional LSTM (Bi-LSTM). Bi-LSTM models are more effective as their output depends on the previous and the next segment, in contrast to the unidirectional LSTM models. The present research aims to map rice crops from Sentinel-1 time series (band C) using LSTM and Bi-LSTM models in West Rio Grande do Sul (Brazil). We compared the results with traditional Machine Learning techniques: Support Vector Machines (SVM), Random Forest (RF), k-Nearest Neighbors (k-NN), and Normal Bayes (NB). The developed methodology can be subdivided into the following steps: (a) acquisition of the Sentinel time series over two years; (b) data pre-processing and minimizing noise from 3D spatial-temporal filters and smoothing with Savitzky-Golay filter; (c) time series classification procedures; (d) accuracy analysis and comparison among the methods. The results show high overall accuracy and Kappa (>97% for all methods and metrics). Bi-LSTM was the best model, presenting statistical differences in the McNemar test with a significance of 0.05. However, LSTM and Traditional Machine Learning models also achieved high accuracy values. The study establishes an adequate methodology for mapping the rice crops in West Rio Grande do Sul.
Fire is one of the primary sources of damages to natural environments globally. Estimates show that approximately 4 million km2 of land burns yearly. Studies have shown that such estimates often ...underestimate the real extent of burnt land, which highlights the need to find better, state-of-the-art methods to detect and classify these areas. This study aimed to analyze the use of deep convolutional Autoencoders in the classification of burnt areas, considering different sample patch sizes. A simple Autoencoder and the U-Net and ResUnet architectures were evaluated. We collected Landsat 8 OLI+ data from three scenes in four consecutive dates to detect the changes specifically in the form of burnt land. The data were sampled according to four different sampling strategies to evaluate possible performance changes related to sampling window sizes. The training stage used two scenes, while the validation stage used the remaining scene. The ground truth change mask was created using the Normalized Burn Ratio (NBR) spectral index through a thresholding approach. The classifications were evaluated according to the F1 index, Kappa index, and mean Intersection over Union (mIoU) value. Results have shown that the U-Net and ResUnet architectures offered the best classifications with average F1, Kappa, and mIoU values of approximately 0.96, representing excellent classification results. We have also verified that a sampling window size of 256 by 256 pixels offered the best results.
Fires associated with the expansion of cattle ranching and agriculture have become a problem in the Amazon biome, causing severe environmental damages. Remote sensing techniques have been widely used ...in fire monitoring on the extensive Amazon forest, but accurate automated fire detection needs improvements. The popular Moderate Resolution Imaging Spectroradiometer (MODIS) MCD64 product still has high omission errors in the region. This research aimed to evaluate MODIS time series spectral indices for mapping burned areas in the municipality of Novo Progresso (State of Pará) and to determine their accuracy in the different types of land use/land cover during the period 2000–2014. The burned area mapping from 8-day composite products, compared the following data: near-infrared (NIR) band; spectral indices (Burnt Area Index (BAIM), Global Environmental Monitoring Index (GEMI), Mid Infrared Burn Index (MIRBI), Normalized Burn Ratio (NBR), variation of Normalized Burn Ratio (NBR2), and Normalized Difference Vegetation Index (NDVI)); and the seasonal difference of spectral indices. Moreover, we compared the time series normalization methods per pixel (zero-mean normalization and Z-score) and the seasonal difference between consecutive years. Threshold-value determination for the fire occurrences was obtained from the comparison of MODIS series with visual image classification of Landsat Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+), and Operational Land Imager (OLI) data using the overall accuracy. The best result considered the following factors: NIR band and zero-mean normalization, obtaining the overall accuracy of 98.99%, commission errors of 32.41%, and omission errors of 31.64%. The proposed method presented better results in burned area detection in the natural fields (Campinarana) with an overall accuracy value of 99.25%, commission errors of 9.71%, and omission errors of 27.60%, as well as pasture, with overall accuracy value of 99.19%, commission errors of 27.60%, and omission errors of 34.76%. Forest areas had a lower accuracy, with an overall accuracy of 98.62%, commission errors of 23.40%, and omission errors of 49.62%. The best performance of the burned area detection in the pastures is relevant because the deforested areas are responsible for more than 70% of fire events. The results of the proposed method were better than the burned area products (MCD45, MCD64, and FIRE-CCI), but still presented limitations in the identification of burn events in the savanna formations and secondary vegetation.
Urbanization processes have caused changes in the runoff behavior, especially by impervious surfaces produced by paving and buildings. Impermeable surfaces prevent the infiltration of rainwater, ...increasing the volume and speed of runoff. Besides, inadequate urban planning coupled with heavy rains promotes the evolution of erosion processes, especially in peri-urban areas. This research aims to identify spatial patterns of geomorphic change in the gully areas due to urbanization in the city of Jacareí (SP). The methodology has the following steps: (1) elaboration of the Digital Elevation Model (DEM) from stereophotogrammetric techniques; (2) elaboration of the pre- and post-urbanization DEM; (3) extraction of contributing area using the D-Infinity method and of the topographic indices (topographic wetness, stream power, and compound topographic); and (4) calculate the difference between the pre- and post-urbanization topographic attributes. The preparation of the pre- and post-urbanization DEM used the MATCH-T DSM and DTMaster modules, both belonging to the INPHO system. Photogrammetric techniques allow the generation of digital models suitable for hydrological studies. The urbanization exposed an evident influence on the triggering of erosion, evidencing an increase of all topographic indices in areas that develop gullies.
The aim in this paper was to evaluate the spatial distribution of protected areas defined by law and their importance as structural corridors. The study area was 7,559,783.69ha located in Western ...Bahia (Northeast Brazil), restricted to the limits of the Urucuia Group (Upper Cretaceous), where there is strong agribusiness growth. Currently, a major dilemma in Brazilian public policy arises from the conflicting interests of environmental conservation and increased food production. Brazilian environmental protection policies include the implantation of Protected Areas (Full Protection Units and Sustainable Use Units) and the adoption of the National Forest Act (Permanent Preservation Areas and Legal Reserves). In this context, we delimited illegal land-use in Permanent Preservation Areas (PPAs) adopting the intersection between land-use/land-cover data from PRISM/ALOS image classification for the years 2007–2010 and PPA vectors. We performed the temporal analysis in PAs considering land-use/land-cover data from Landsat TM image classification for the years 1988, 1992, 1996, 2000, 2004, 2008 and 2011. Finally, we performed a Morphological Spatial Pattern Analysis (MSPA) to evaluate whether PPAs alone are sufficient as structural corridors. Hypothetical scenarios were simulated to increase the potential of PPAs as structural corridors. The calculation of MSPA attributes was conducted considering 3 edge widths: 15m (1 pixel), 510m (34 pixels), and 1005m (67 pixels). Four scenarios were simulated, considering a gradual increase in preserved areas. The results show that illegal land use is contained within PPA and protected areas. The scenario simulations present alternatives to increase the connectivity of the fragments and ensure the maintenance of ecological and hydrological services. Rapid agricultural expansion without proper landscape planning can compromise the sustainability of ecosystem services and the recharge zone of the Urucuia aquifer.
The state of Amapá within the Amazon biome has a high complexity of ecosystems formed by forests, savannas, seasonally flooded vegetation, mangroves, and different land uses. The present research ...aimed to map the vegetation from the phenological behavior of the Sentinel-1 time series, which has the advantage of not having atmospheric interference and cloud cover. Furthermore, the study compared three different sets of images (vertical–vertical co-polarization (VV) only, vertical–horizontal cross-polarization (VH) only, and both VV and VH) and different classifiers based on deep learning (long short-term memory (LSTM), Bidirectional LSTM (Bi-LSTM), Gated Recurrent Units (GRU), Bidirectional GRU (Bi-GRU)) and machine learning (Random Forest, Extreme Gradient Boosting (XGBoost), k-Nearest Neighbors, Support Vector Machines (SVMs), and Multilayer Perceptron). The time series englobed four years (2017–2020) with a 12-day revisit, totaling 122 images for each VV and VH polarization. The methodology presented the following steps: image pre-processing, temporal filtering using the Savitsky–Golay smoothing method, collection of samples considering 17 classes, classification using different methods and polarization datasets, and accuracy analysis. The combinations of the VV and VH pooled dataset with the Bidirectional Recurrent Neuron Networks methods led to the greatest F1 scores, Bi-GRU (93.53) and Bi-LSTM (93.29), followed by the other deep learning methods, GRU (93.30) and LSTM (93.15). Among machine learning, the two methods with the highest F1-score values were SVM (92.18) and XGBoost (91.98). Therefore, phenological variations based on long Synthetic Aperture Radar (SAR) time series allow the detailed representation of land cover/land use and water dynamics.
Instance segmentation is the state-of-the-art in object detection, and there are numerous applications in remote sensing data where these algorithms can produce significant results. Nevertheless, one ...of the main problems is that most algorithms use Red, Green, and Blue (RGB) images, whereas Satellite images often present more channels that can be crucial to improve performance. Therefore, the present work brings three contributions: (a) conversion system from ground truth polygon data into the Creating Common Object in Context (COCO) annotation format; (b) Detectron2 software source code adaptation and application on multi-channel imagery; and (c) large scene image mosaicking. We applied the procedure in a Center Pivot Irrigation System (CPIS) dataset with ground truth produced by the Brazilian National Water Agency (ANA) and Landsat-8 Operational Land Imager (OLI) imagery (7 channels with 30-m resolution). Center pivots are a modern irrigation system technique with massive growth potential in Brazil and other world areas. The round shapes with different textures, colors, and spectral behaviors make it appropriate to use Deep Learning instance segmentation. We trained the model using 512 × 512-pixel sized patches using seven different backbone structures (ResNet50- Feature Pyramid Network (FPN), Resnet50-DC5, ResNet50-C4, Resnet101-FPN, Resnet101-DC5, ResNet101-FPN, and ResNeXt101-FPN). The model evaluation used standard COCO metrics (Average Precision (AP), AP50, AP75, APsmall, APmedium, and AR100). ResNeXt101-FPN had the best results, with a 3% advantage over the second-best model (ResNet101-FPN). We also compared the ResNeXt101-FPN model in the seven-channel and RGB imagery, where the multi-channel model had a 3% advantage, demonstrating great improvement using a larger number of channels. This research is also the first with a mosaicking algorithm using instance segmentation models, where we tested in a 1536 × 1536-pixel image using a non-max suppression sorted by area method. The proposed methodology is innovative and suitable for many other remote sensing problems and medical imagery that often present more channels.
Vehicle classification is a hot computer vision topic, with studies ranging from ground-view to top-view imagery. Top-view images allow understanding city patterns, traffic management, among others. ...However, there are some difficulties for pixel-wise classification: most vehicle classification studies use object detection methods, and most publicly available datasets are designed for this task, creating instance segmentation datasets is laborious, and traditional instance segmentation methods underperform on this task since the objects are small. Thus, the present research objectives are as follows: first, propose a novel semisupervised iterative learning approach using the geographic information system software, second, propose a box-free instance segmentation approach, and third, provide a city-scale vehicle dataset. The iterative learning procedure considered the following: first, labeling a few vehicles from the entire scene, second, choosing training samples near those areas, third, training the deep learning model (U-net with efficient-net-B7 backbone), fourth, classifying the whole scene, fifth, converting the predictions into shapefile, sixth, correcting areas with wrong predictions, seventh, including them in the training data, eighth repeating until results are satisfactory. We considered vehicle interior and borders to separate instances using a semantic segmentation model. When removing the borders, the vehicle interior becomes isolated, allowing for unique object identification. Our procedure is very efficient and accurate for generating data iteratively, which resulted in 122 567 mapped vehicles. Metrics-wise, our method presented higher intersection over union when compared to box-based methods (82% against 72%), and per-object metrics surpassed 90% for precision and recall.
LiDAR is an excellent source of elevation data used in many surveys. The spaceborne handle system, Global Ecosystem Dynamics Investigation (GEDI), provides ground elevation information with high ...accuracy except for areas with steep slopes. GEDI data have a lot of noise from atmospheric conditions, and therefore filtering procedures are mandatory to select the best dataset. The dataset presents uncertainties of different magnitudes, with values reaching more than 100 m of difference between the reference data and the GEDI data. The challenge is to find a criterion to determine a threshold to filter accurate GEDI samples. This research aims to identify the threshold based on the difference values between the reference data and the GEDI data to select the maximum number of samples with low RMSE values. Therefore, we used the Kolmogorov–Smirnov (KS) non-parametric test to define the best threshold based on a normal distribution. Our results demonstrated a lower RMSE value with a high number of samples when compared with the quality flag parameter threshold, even using sensitivity parameter thresholds. This method is useful for achieving the best possible accuracy from GEDI data worldwide.
PDF
