National Institute for Space Research
Biblioteca Digital da Memória Científica do INPENot a member yet
6175 research outputs found
Sort by
Avaliação do NDVI de acordo com as fitofisionomias e tipos de solo na bacia hidrográfica do Rio São Marcos/GO-MG
No full textThe use of remote sensing data and techniques allow land cover dynamics studies including vegetation phenology. The aim of this study was to evaluate the Normalized Difference Vegetation Index (NDVI) in the São Marcos River basin, with the purpose of analysing the dynamics of cerrado vegetation typologies, considering different soil types. For this, the following steps were performed: (1) combination of vegetation (PROBIO) and soil (MMA/IBAMA) maps using SPRING system functions; (2) vectorization of eighteen classes after the vegetation/soil integration; (3) extraction of averaged NDVI for each of the classes in forty-six MODIS NDVI scenes, for the years 2001 and 2011; and (4) NDVI values assembling and ploting. The results showed a similar seasonal distribution of NDVI values for the classes of grazing/pasture, savanna park and woody-grass savanna, showing mainly the difference between rainy and dry season, in all soil types. Agriculture was the land use/class that presented the greatest difference in NDVI between the summer and winter months, explained by practice of monoculture (periods with crop cultivation and bare soil). Considering the variation between the two years studied, the highest NDVI values occurred for the year of 2011 for all vegetation typologies analysed.Pages: 2755-276
Morfologia de fundo do litoral do Camocim Estado do Ceará
No full textThe continental shelf of the Ceará State is characterized by having low gradient slope and large homogeneous areas disrupted by paleo-river valleys, including Coreaú river. To achieve the proposed objective were used Bathymetry Date and satellite image interpretation. For making mapping were used scientifically techniques recognized. We identified and characterized six different geomorphological region in the paleo-canyon. This work is part of project "Potential and Environmental Management in Granular Exploration of the Continental Shelf of the State of Ceará.Pages: 7848-785
Identificação e Análise das Áreas de Proteção Permanente no Médio Curso da Bacia do Rio Piraqueaçú/ES com o Uso de Geotecnologias
No full textThis work, which is part of a larger project, aims to identify and analyze the main Permanent Preservation Areas (PPAs) of basin Piraqueaçu river, northern Espírito Santo state - Brazil based on the new legislation. Initially readings were made in books, theses, dissertations, monographs, articles, relevant legislation in order to substantiate the article, followed the acquisition cartographic as orthophotos and Plans of Information. The results were satisfactory enabling identification of PPAs in some sectors of the watercourses, predominantly in over 55% of cases, the activity / land use, irregularly.Pages: 4287-429
Análise de dados do satélite WorldView-2 baseada em objetos, para a classificação das áreas de manguezais em São Luis (MA), Brasil
No full textThe objective of this paper is to classify different types of irregular occupation in areas covered by mangroves in São Luis (Maranhão State, Brazil), considering the paradigm OBIA (Object-based Image Analysis), analyzing WorldView-2 satellite data with InterIMAGE, a free and open source software for image analysis. A methodology was proposed to study the main targets of this area, such as: secondary vegetation (apicum), mangroves, tidal flats, irregular constructions in tidal areas, paved streets and different types of condominiums. Initially an image fusion was made, merging multispectral images with the panchromatic channel to improve information quality of WorldView-2 data. Following an ortho-rectification was made, using GCPs obtained during field survey. The data mining GEODMA software was used to establish the attributes which characterize the targets of interest. Afterwards those classes of interest were structured and a knowledge model was established. The classification using the OBIA paradigm allowed the discrimination of mangrove areas as well as those sections occupied irregularly. The use of open source free software InterIMAGE (for image classification) and GEODMA for the elaboration of a decision tree allowed a fast classification, from optimized image attributes. The additional new channels Yellow and Red Edge at the WorldView-2 sensor system were extremely important to discriminate the mangrove from other vegetation types.Pages: 5986-599
A influência do elemento estruturante no processo de detecção de nuvens em imagens de sensoriamento remoto
No full textThe broad applicability of Digital Image Processing (DIP) has generated many scientific effort in this area. Associated to Remote Sensing, the DIP area gained significant importance with regard to the study and analysis of large regions, however, images acquired by sensors are subject to interference and noise. Due to this, scientific efforts are being employed in order to eradicate or at least minimize the noise in multispectral images. This paper focus on the preprocessing stage of DIP in remote sensing, which is proposed a generalized structuring elements size for TerraLib library (open source library for working with Geographic Information System (GIS) developed in partnership with the National Institute for Space Research (INPE), with the Computer Graphics Technology (Tecgraf /PUC - Rio) and the Foundation for the Space Science, Applied Research and Technology (FUNCATE)), proposing greater freedom to detect regions of interest in images. It is applied a morphological opening using primitive processing operations such as erosion and dilation in noisy images (affected by regions of sparse or dense cloud and shadow) to comparative analysis of the results obtained. Thus are observed satisfactory results in images of different resolutions as the freedom to choose the shape and size of the structuring element applied, detecting regions of interest (clouds and shadows) in these images. This stage contributes in noise detection in order to eliminate them through DIP techniques, ensuring more refined analysis in subsequent stages of preprocessing, such as segmentation and classification.Pages: 8011-801
Cultural Heritage Surveying: The Experiences Of The Photogrammetry Laboratory Of The Politecnico Di Bary In Summer School 2011
No full textThis paper reports the experiences of the second edition of an international, educational programme named Summer School 2011. It was held by the Laboratory of Architectonic Photogrammetry of the Politecnico di Bari Institute, in Italy. The main goals were to spread the importance of cultural heritage preservation, and to emphasize the potential of application of LASER scanning techniques to the surveying of cultural heritage sites. It involved five professors and seventeen professionals and students from different areas from Italy, Brazil, Turquey, Venezuela, and Romania. It had an itinerant programme, from the city of Noci, to the community of Laterza, and the cities of Lecce and Bríndisi. The methodology adopted was: (a) technical presentations; (b) the on-site presentations and descriptions of the cultural and historical highlights of the objects of interest, and; (c) the demonstrations and practice of LASER scanning surveying. This paper also reports the results of the surveying at The Cantina Spagnola cavern, and The Tiempio di S. Giovanni al Sepolcro in Bríndisi, Italy. The main conclusion was the spreading of the thinking about cultural heritage inventory techniques, relevant for maintaining the historical records of a country. The Laboratory of photogrammetry of The Rio de Janeiro State University, in Brazil, has started a joint technical project with the Politecnico di Bari Institute toward the surveying of two historical sites in the State of Rio de Janeiro: (a) the ruins of a city, São João Marcos, and; (b) the Palácio Guanabara. This paper includes the results of that international cooperation.Pages: 6156-616
Potencial das imagens de satélites gratuitas para o mapeamento da bacia do rio Ocoy - oeste do PR
No full textGeotechnologies are excellent tools if applied to the development of studies that involve the environment. Being widely available on the Internet, they become available to all educational and research institutions. The proposal of this work is thus to show a routine of development of thematic maps (slope, hypsometry and land use maps) of the Ocoy river, generated using the free software SPRING and free images from the LANDSAT 5 satellite, besides images from the SRTM mission. This allows for a wide vision of how to use those resources to generate information about the environment. The Ocoy river is a potential object of research because it is an affluent of the Paraná River, and because its basin is mostly comprised by agricultural land and some urban sites, such as Medianeira and Matelândia, responsible for the formation of industrial and urban waste found in the river. Results found in the survey of land use of the Ocoy basin with free technologies of Remote Sensing and Geographic Information Systems have shown to be satisfactory, enabling an easy interpretation of the data and an integrated and interdisciplinary analysis for planning and integrated management of the exploitation of the available resources.Pages: 5243-525
Análise Comparativa da Classificação Espectro-Textural de Imagens Fusionadas CCD e HRC/CBERS-2B Através de Redes Neurais Artificiais
No full textHRC and CCD / CBERS-2B data fusion provides improvements in the textural images, which have been little explored in automatic classification. In this work, the thematic accuracies of Land Use / Land Cover mapping obtained from neural networks like MLP (Multilayer Perceptrons) were performed through two methodologies: architectures with input windows ranging from 3x3 to 9x9 versus MLP networks with simple inputs. For the second architecture, the input vector consisted of three textural descriptors (Contrast, Entropy and Second Angular Momentum) derivatives of windows ranging from 3x3 to 9x9, plus data from the red and infrared bands. Through the Kappa statistics was found that the input windows should have no dimension larger than 5x5. However, the use of networks with textural descriptors improved the accuracies when the size of the windows were increased, presenting better performance compared to the MLP networks with inputs in NxN windows.Pages: 2234-224
Utilizando ferramenta SIG para subsídio ao inventário de um produto florestal não madeireiro (PFNM) do tipo cipó, em área florestal Amazônica
No full textThe vine-titica is considered a non-timber forest product (NTFP), typical of the Amazon rainforest land. High economic value for its fiber is characteristically light in color, flexible, durable and good workability. There is little information regarding the best methods for extraction and handling of this vine, and the NTFP in general. For inventory of NTFPs there is not a methodology for each type of product, and what you have are arboreal adaptations inventory and zoo. It becomes important studies aimed at the applicability of different inventory methods, providing greater flexibility and economy to survey these products in the forest. Thus, this study aims to define the size and shape of plots, with a degree of efficiency and lower cost for the inventory of a type of vine NTFP, using simulations in a powerful GIS. All hosts of vine-titica inventoried were plotted by the coordinates collected in the field and within the 12 plots (1.5 ha each). To assess the optimum size and shape of plots for the inventory of vine-titica, simulations with 14 different sizes and shapes of plots were made using the ArcGIS 9.3. The number of samples for each category of size and shape of portion was 36 repetitions. With the data obtained from the simulations, we compared the different sizes and shapes of plots through the density of host trees per hectare per share, the number of vines per hectare per share, uncertainty and economic comparison (proposed by Cochran). We used analysis of variance (ANOVA) to test the significance and Dunnett's test. The sizes and shapes of plots of 50x50 m, m 25x100, 20x150 50x60m had lower relative variance or lower relative cost.Pages: 120-12
Simulação de Monte Carlo na avaliação das incertezas em parte do processo de calibração de sensores
No full textThe airborne or satellite sensors absolute calibration is based on the relationship between Digital Numbers and top-of-atmosphere (TOA) radiance from a reference surface located on the Earth. This reference surface has been spectrally characterized by radiometers or spectrometers. The Digital Number generated by the sensor in each spectral band depends not only on the amount of incident energy, but also on the spectral response function at each band. Therefore, before comparing the Digital Number from the orbital imagery to the spectral data collected from spectroradiometers it is necessary to weigh them taking into account the spectral response function of each sensor band. To calculate the uncertainties of both the measurement process and weighting it was applied the Monte Carlo simulation. This method uses a random number sequence to perform a simulation. The TM sensor onboard of the Landsat 5 satellite was used to test the proposed methodology. Results showed that the standard uncertainty for each sensor bands TM ranging from 0.30% to 0.93%.Pages: 9131-913