Organização e armazenamento de imagens multitemporais georreferenciadas para suporte ao processo de detecção de mudanças
Carregando...
Arquivos
Data
2018-03-01
Autores
Souza, Luiz Eduardo Christovam de [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Atualmente o volume de dados produzidos tem atingido patamares nunca imaginados, sobretudo em decorrência da multiplicação do número de sensores e da popularização da internet, com a web 2.0 e as redes sociais. Dentre os diversos tipos de sensores existentes, os de imageamento, transportados principalmente por satélites, produzem vastos conjuntos de observações da superfície da Terra. A observação contínua da Terra por satélites possibilita o monitoramento de mudanças no uso e cobertura da terra. Contudo, em diversas pesquisas relacionadas a mudanças no planeta, são utilizados apenas pequenos fragmentos do imenso conjunto de dados existente, essencialmente devido a ainda haver uma lacuna científicatecnológica relacionada aos procedimentos de organização, armazenamento, análise e representação de grandes conjuntos de dados. Portanto, nessa pesquisa foi definida uma estrutura para organização, armazenamento e recuperação de dados espaço-temporais, com o propósito de fornecer suporte a detecção de mudanças na cobertura da terra. Para tanto, foi definida como aplicação a análise de séries temporais de Normalized Difference Vegetation Index (NDVI) derivadas de imagens adquiridas desde 1984 até 2017, pelos sensores Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+) e Operational Land Imager (OLI) para a região de Porto Velho, Rondônia. Foi construída uma série temporal de NDVI para a posição de cada pixel presente na área de estudo. Regiões de referência foram definidas para recuperar séries temporais de referência que descrevem os tipos de cobertura da terra e classes de mudanças (antrópicas ou naturais). O algoritmo Fast Dynamic Time Warping (FastDTW), foi adotado para medir a similaridade entre séries temporais, a serem classificadas e de referência. Para determinar a classe da série temporal, as distâncias obtidas com o FastDTW foram classificadas com o algoritmo k-Nearest Neigbor (kNN). A classificação de várias séries permitiu a construção do mapa de mudanças para uma área teste inserida na área de estudo. A respeito do banco de dados, esse foi definido com uma abordagem híbrida, foram utilizados o modelo relacional, que permitiu armazenar dados rigidamente estruturados, e o modelo de dados não relacional (NoSQL), que possibilitou armazenar dados semiestruturados. Essa estrutura de armazenamento foi fundamental para a automatização da aplicação, permitindo a organização, armazenamento e recuperação de um extenso conjunto de dados heterogêneos de forma consistente e segura. Com relação às séries temporais foi possível observar que de maneira geral o perfil temporal de NDVI pode descrever o tipo de cobertura da terra, entretanto efeitos atmosféricos, nuvens, fumaça, entre outros, exercem influência no comportamento da série temporal. Foram realizadas duas classificações, com tamanhos de raio iguais a 1 e 20 para o FastDTW. A concordância global dos mapas produzidos foi de 58,06% e 55,02%, respectivamente. Apesar desses valores serem relativamente baixos, esses resultados são justificados pelo caráter subjetivo presente na caracterização de mudanças temporais, pela não filtragem das séries temporais e pela elevada variação no tamanho das séries.
Nowadays the size of datasets has been reaching levels never seen before, mainly due to new sensors and the widespread of the internet, with web 2.0 and social media. Among the various types of sensors, the imaging sensors, mainly carried by satellites, have produced big Earth observations datasets. The regular Earth observation by satellites enable to monitor Land Use/Cover Change (LUCC). However, in many researches related to LUCC, only small parts of the big Earth Observation datasets are normally used, because there is still a scientifictechnological gap related to the organization, storage, analysis and representation of big Earth Observations data. Therefore, in this research was defined a database for the organization, storage and retrieval of spatio-temporal data, to support a LUCC task. Therefore, the time series analysis of Normalized Difference Vegetation Index (NDVI) of images acquired from 1984 to 2017 by Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+) and Operational Land Imager (OLI) for the region of Porto Velho, Rondônia was defined as the application. To the position each of the pixel in the study area was built a NDVI time series. Reference areas were defined to retrieve reference time series that describe the land cover types and the change classes (anthropic and natural). The Fast Dynamic Time Warping (FastDTW) algorithm was used to measure the similarity between the time series, to be classified and reference ones. To find the time series class, the distances previously found by FastDTW were classified by kNearest Neigbor (kNN) algorithm. Classifying several time series allowed the construction of LUCC map in a test area. Regarding the database, this was defined with a hybrid approach, the relational data model allowed to store highly structured data, and the non-relational data model (NoSQL) allowed store semi-structured data. The database was important in automating the application, enabling the organization, storage and retrieval of an extensive heterogeneous dataset in a consistent and secure way. Regarding the time series, it was possible to observe that in general the NDVI time series can describe land cover types, however atmospheric effects, clouds, smoke, among others, can influence the way of the time series. Two classifications were performed using the FastDTW, the first one with radio size equal to 1 and for the second the radio was 20. The overall accuracy of the maps was 58.06% and 55.02%, respectively. Although these are relatively low values, they are justified by the subjective presented in the temporal changes characterization, by the use of the raw time series and by the high variation in the time series size.
Nowadays the size of datasets has been reaching levels never seen before, mainly due to new sensors and the widespread of the internet, with web 2.0 and social media. Among the various types of sensors, the imaging sensors, mainly carried by satellites, have produced big Earth observations datasets. The regular Earth observation by satellites enable to monitor Land Use/Cover Change (LUCC). However, in many researches related to LUCC, only small parts of the big Earth Observation datasets are normally used, because there is still a scientifictechnological gap related to the organization, storage, analysis and representation of big Earth Observations data. Therefore, in this research was defined a database for the organization, storage and retrieval of spatio-temporal data, to support a LUCC task. Therefore, the time series analysis of Normalized Difference Vegetation Index (NDVI) of images acquired from 1984 to 2017 by Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETM+) and Operational Land Imager (OLI) for the region of Porto Velho, Rondônia was defined as the application. To the position each of the pixel in the study area was built a NDVI time series. Reference areas were defined to retrieve reference time series that describe the land cover types and the change classes (anthropic and natural). The Fast Dynamic Time Warping (FastDTW) algorithm was used to measure the similarity between the time series, to be classified and reference ones. To find the time series class, the distances previously found by FastDTW were classified by kNearest Neigbor (kNN) algorithm. Classifying several time series allowed the construction of LUCC map in a test area. Regarding the database, this was defined with a hybrid approach, the relational data model allowed to store highly structured data, and the non-relational data model (NoSQL) allowed store semi-structured data. The database was important in automating the application, enabling the organization, storage and retrieval of an extensive heterogeneous dataset in a consistent and secure way. Regarding the time series, it was possible to observe that in general the NDVI time series can describe land cover types, however atmospheric effects, clouds, smoke, among others, can influence the way of the time series. Two classifications were performed using the FastDTW, the first one with radio size equal to 1 and for the second the radio was 20. The overall accuracy of the maps was 58.06% and 55.02%, respectively. Although these are relatively low values, they are justified by the subjective presented in the temporal changes characterization, by the use of the raw time series and by the high variation in the time series size.
Descrição
Palavras-chave
Banco de dados híbrido, NoSQL, Séries temporais de imagens, LUCC, DTW, Hybrid database, Remote sensing image time series