Mapping deforestation is an essential step in the process of managing tropical rainforests. It lets us understand and monitor both legal illegal its implications, which include effect may have on climate change through greenhouse gas emissions. Given that there ample room for improvements when it comes to mapping using satellite imagery, this study, we aimed test evaluate use algorithms belonging growing field deep learning (DL), particularly convolutional neural networks (CNNs), end....

The Synthetic Aperture Radar (SAR) time series allows describing the rice phenological cycle by backscattering signature. Therefore, advent of Copernicus Sentinel-1 program expands studies radar data (C-band) for monitoring at regional scales, due to high temporal resolution and free distribution. Recurrent Neural Network (RNN) model has reached state-of-the-art in pattern recognition time-sequenced data, obtaining a significant advantage crop classification on remote sensing images. One...

The need to monitor the Earth’s surface over a range of spatial and temporal scales is fundamental in ecosystems planning management. Change-Vector Analysis (CVA) bi-temporal method change detection that considers magnitude direction vector. However, many multispectral applications do not make use component. procedure most used calculate component using multiband data cosine, but number output cosine images equal original bands has complex interpretation. This paper proposes new approach...

Movement of livestock production within a country or region has implications for genetics, adaptation, well-being, nutrition, and logistics, particularly in continental-sized countries, such as Brazil. Cattle Brazil from 1977 to 2011 was spatialized, the annual midpoint calculated. Changes relative acceleration were calculated spatialized using ARCGIS®. Cluster canonical discriminant analyses performed further highlight differences between regions terms cattle production. The mean point...

Instance segmentation is the state-of-the-art in object detection, and there are numerous applications remote sensing data where these algorithms can produce significant results. Nevertheless, one of main problems that most use Red, Green, Blue (RGB) images, whereas Satellite images often present more channels be crucial to improve performance. Therefore, work brings three contributions: (a) conversion system from ground truth polygon into Creating Common Object Context (COCO) annotation...

Brazil is a tropical country with continental dimensions and abundant solar resources that are still underutilized. However, energy one of the most promising renewable sources in country. The proper inspection Photovoltaic (PV) plants an issue great interest for Brazilian territory’s management agency, advances computer vision deep learning allow automatic, periodic, low-cost monitoring. present research aims to identify PV using semantic segmentation mosaicking approach large image...

Remote sensing has been used in karst studies to identify limestone terrain, describe exokarst features, analyze depressions, and detect geological structures important development. The aim of this work is investigate the use ASTER-, SRTM- ALOS/PRISM-derived digital elevation models (DEMs) quantify natural depressions along São Francisco River near Barreiras city, northeast Brazil. study area a landscape characterized by (dolines), closed limestone, many which contain standing water...

Panoptic segmentation combines instance and semantic predictions, allowing the detection of "things" "stuff" simultaneously. Effectively approaching panoptic in remotely sensed data can be auspicious many challenging problems since it allows continuous mapping specific target counting. Several difficulties have prevented growth this task remote sensing: (a) most algorithms are designed for traditional images, (b) image labelling must encompass classes, (c) annotation format is complex. Thus,...

Oil spills are a worldwide concern since they cause environmental problems and financial losses. Automatic detection plays crucial role in rapid decision-making to reduce damage. In this context, deep learning remote sensing powerful tools with successful applications many regions. However, there still no studies on deep-water zones the Brazilian territory. The present research has three contributions: (1) create an oil spill dataset region, (2) compare state-of-the-art models for task, (3)...

Mass movements in Brazil are common phenomena, especially during strong rainfall events that occur frequently the summer season. These phenomena cause losses of lives and serious damage to roads, bridges, properties. Moreover, illegal occupation by slums on slopes around cities intensifies effect mass movement. This study aimed develop a methodology combines models shallow landslides debris-flows order create map with initiation volume runout distance. The area comprised two catchments Rio...

Radiometric precision is difficult to maintain in orbital images due several factors (atmospheric conditions, Earth-sun distance, detector calibration, illumination, and viewing angles). These unwanted effects must be removed for radiometric consistency among temporal images, leaving only land-leaving radiances, optimum change detection. A variety of relative correction techniques were developed the or rectification same area, through use reference targets whose reflectance do not...

Predicting the spatial distribution of wildfires is an important step towards proper wildfire management. In this work, we applied two data-mining models commonly used to predict fire occurrence – logistic regression (LR) and artificial neural network (ANN) Brazil’s Federal District, located inside Brazilian Cerrado. We Landsat-based burned area products generate dependent variable, nine different anthropogenic environmental factors as explanatory variables. The were optimised via feature...

The Cerrado is the second largest biome in Brazil after Amazon and savanna with highest biodiversity world. Serra Tombador Natural Reserve (STNR) private reserve located Goiás State, fourth biome. present study aimed to map burnt areas describe spatial patterns of fire recurrence its interactions classes land-cover that occurred STNR surroundings period between 2001 2010. Several Landsat TM images acquired around months July, August September, coinciding region’s dry season when events...

The center pivot irrigation system (CPIS) is a modern technique widely used in precision agriculture due to its high efficiency water consumption and low labor compared traditional methods. CPIS leader mechanized Brazil, with growth forecast for the coming years. Therefore, mapping of areas strategic factor estimation agricultural production, ensuring food security, resources management, environmental conservation. In this regard, digital processing satellite images primary tool allowing...

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 use object detection methods, and publicly available datasets designed this task, creating instance segmentation laborious, traditional methods underperform on task since the objects small. Thus, present research...

Atualmente, os alagamentos, embora sejam eventos curtos e rápidos, tendem a ser mais frequentes. O Distrito Federal é acometido todos anos por causando uma diversidade de impactos negativos como, danos materiais, ambientais perdas vidas. proposta do artigo atualizar base dados (não oficial) demonstrar as possibilidades uso colaborativos. Como resultado, foi observado que no período atualização (04/2013 até 12/2020) houve um incremento 50% das notificações em relação ao previamente cadastrado...

Sheep production is present on all continents and has been practiced in Brazil since the colonization. In this study, multitemporal dynamics of sheep examined using official government data (Brazilian Institute for Geography Statistics-IBGE) from 1976 to 2010. Maps flock growth rates acceleration maps by municipality were elaborated. The Southern states are seen show a reduction mainly due wool crisis 1970s 80s. Northeast be important meat production. More recently, centerwest northern have...

We have mapped the primary native and exotic vegetation that occurs in Cerrado-Caatinga transition zone Central Brazil using MODIS-NDVI time series (product MOD09Q1) data over a two-year period (2011–2013). Our methodology consists of following steps: (a) development three-dimensional cube composed NDVI-MODIS series; (b) removal noise; (c) selection reference temporal curves classification similarity distance measures; (d) support vector machines (SVMs). evaluated different classifications...

The expansion of agricultural frontiers in Brazil has caused substantial changes land use and cover. This research aims to analyze the space-time dynamics soybeans cattle production Brazilian territory during period 1991–2015. spatial analysis adopted following procedures: (a) change vector from annual calculation midpoint production; (b) mapping growth acceleration rates two productions, (c) correlation between time series soybean cattle. results showed high for soy South, Central-West...

Fire is one of the primary sources damages to natural environments globally. Estimates show that approximately 4 million km2 land burns yearly. Studies have shown such estimates often underestimate real extent burnt land, which highlights need find better, state-of-the-art methods detect and classify these areas. This study aimed analyze use deep convolutional Autoencoders in classification areas, considering different sample patch sizes. A simple Autoencoder U-Net ResUnet architectures were...