1,720,969 research outputs found
Développement d'une nouvelle approche basée objets pour l'extraction automatique de l'information géographique en milieu urbain à partir des images satellitaires à très haute résolution spatiale
No full textRésumé: L'importance de l'information géographique est indéniable pour des prises de décision efficaces dans le milieu urbain. Toutefois, sa disponibilité n'est pas toujours évidente. Les images satellitaires à très résolution spatiale (THRS) constituent une source intéressante pour l'acquisition de ces informations. Cependant, l'extraction de l'information géographique à partir de ces images reste encore problématique. Elle fait face, d'une part, aux spécificités du milieu urbain et celles des images à THRS et d'autre part, au manque de méthodes d'analyse d'images adéquates. Le but de la présente étude est de développer une nouvelle approche basée objets pour l'extraction automatique de l'information géographique en milieu urbain à partir des images à THRS. L'approche proposée repose sur une analyse d'image basée objets. Deux étapes principales sont identifiées : le passage des pixels aux primitives objets et le passage des primitives aux objets finaux. La première étape est assurée par une nouvelle approche de segmentation multispectrale non paramétrée. Elle se base sur la coopération entre les segmentations par régions et par contours. Elle utilise un critère d'homogénéité spectrale dont le seuil est déterminé d'une manière adaptive et automatique. Le deuxième passage part des primitives objets créées par segmentation. Elle utilise une base de règles floues qui traduisent la connaissance humaine utilisée pour l'interprétation des images. Elles se basent sur les propriétés des objets des classes étudiées. Des connaissances de divers types sont prises en considération (spectrales, texturales, géométriques, contextuelles). Les classes concernées sont : arbre, pelouse, sol nu et eau pour les classes naturelles et bâtiment, route, lot de stationnement pour les classes anthropiques. Des concepts de la théorie de la logique floue et celle des possibilités sont intégrés dans le processus d'extraction. Ils ont permis de gérer la complexité du sujet étudié, de raisonner avec des connaissances imprécises et d'informer sur la précision et la certitude des objets extraits. L'approche basée objets proposée a été appliquée sur des extraits d'images Ikonos et Quickbird. Un taux global de 80 % a été observé. Les taux de bonne extraction trouvés pour les classes bâtiment, route et lots de stationnement sont de l'ordre de 81 %, 75 % et 60 % respectivement. Les résultats atteints sont intéressants du moment que la même base des règles a été utilisée. L'aspect original réside dans le fait que son fonctionnement est totalement automatique et qu'elle ne nécessite ni données auxiliaires ni zones d'entraînement. Tout le long des différentes étapes de l'approche, les paramètres et les seuils nécessaires sont déterminés de manière automatique. L'approche peut être transposable sur d'autres sites d'étude. L'approche proposée dans le cadre de ce travail constitue une solution intéressante pour l'extraction automatique de l'information géographique à partir des images à THRS.||Abstract: The importance of the geographical information is incontestable for efficient decision making in urban environment. But, it is not always available.The very high spatial resolution (VHSR) satellite images constitute an interesting source of this information. However, the extraction of the geographical information from these images is until now problematic.The goal of the present study is to develop a new object-based approach for automatic extraction of geographical information in urban environment from very high spatial resolution images.The proposed approach is object-based image analysis. There are two principal steps: passage of pixels to object primitives and passage of primitives to final objects.The first stage uses a new multispectrale cooperative segmentation approach. Cooperation between region and edge information is exploited. Segments are created with respect to their spectral homogeneity.The threshold is adaptive and its determination is automatic.The second passage leaves from object primitives created by segmentation. Fuzzy rule base is generated from the human knowledge used for image interpretation. Several kinds of object proprieties are integrated (spectral, textural, geometric, and contextual).The concerned classes are trees, grass, bare soil and water as natural classes and building, road, parking lot as man made classes. Fuzzy logic and possibilities theories are integrated in the process of extraction. They permitted to manage the complexity of the studied objects, to reason with imprecise knowledge and to inform on precision and certainty of the extracted objects.The approach has been applied with success on various subsets of Ikonos and Quickbird images.The global extraction accuracy was about 80%.The object-based approach was able to extract buildings, roads and parking lots in urban areas with of 81%, 75% and 60% extraction accuracies respectively.The results are interesting with regard to that the same rule base was used.The original aspect resides in the fact that the approach is completely automatic and no auxiliary data or training areas are required. Along the different stages of the approach, the parameters and the thresholds are determined automatically. This allows the transposability of the approach on others VHRS images.The present approach constitutes an interesting solution for automatic extraction of the geographical information from VHSR satellite images
Random forest effectiveness for Bragança region mapping: comparing indices, number of the decision trees, and generalization
Get PDFMestrado de dupla diplomação com o Institute Agronomic and Veterinary Hassan IIRemote sensing is a domain that tends to use satellite images for classification and Land Use/Cover (LULC) mapping. For this purpose, classification algorithms are used, which are numerous and diverse, and it is necessary to establish decision criteria when choosing the algorithm. Ultimately, the main decision criterion will be the accuracy obtained in classification because the accuracy of classification may differ from one algorithm to another, even within the same algorithm, according to its variables. But there are other equally important criteria: it depends on the nature of the task, the quantity and types of data available, the type of response expected, the time and computational resources available, the depth of our knowledge about the algorithms.
The methodology of each part of the work was described and the criteria for comparison were established. In this research, with the same training data, the same validation data, the same application context (7 classes), and the same image data (Sentinel-2), we tested 15 iterations with the Random Forest classification algorithm, with different tree number decision values, and 3 iterations with vegetation and soil indexes, for the production of the LULC map of the Bragança region (northeast Portugal). Finally, we evaluate the accuracy of the classification, before and after the post-classification tasks (generalization, fragmentation and removal of isolated pixels).
The results obtained show that a classification with an nb-trees = 1000, including vegetation and soil indices, and after post-classification tasks, provided excellent precision results (Coefficient Kappa = 0.93, Overall accuracy = 96%, and marginal errors of omission & commission below 4%).A teledetecção é um domínio que tende a utilizar imagens de satélite para classificação e mapeamento de Uso/Cobertura da Terra (LULC). Para este fim, são utilizados algoritmos de classificação, que são numerosos e diversos, sendo necessário estabelecer critérios de decisão ao escolher o algoritmo. Em última análise, o principal critério de decisão será a precisão obtida na classificação, porque a precisão da classificação pode diferir de um algoritmo para outro, mesmo dentro do mesmo algoritmo, de acordo com as suas variáveis. Mas existem outros critérios igualmente importantes: depende da natureza da tarefa, da quantidade e tipos de dados disponíveis, do tipo de resposta esperada, do tempo e dos recursos computacionais disponíveis, da profundidade dos nossos conhecimentos sobre os algoritmos.
A metodologia de cada parte do trabalho foi descrita e os critérios de comparação foram estabelecidos. Nesta investigação, com os mesmos dados de formação, os mesmos dados de validação, o mesmo contexto de aplicação (7 classes), e os mesmos dados de imagem (Sentinel-2), testámos 15 iterações com o algoritmo de classificação Random Forest, com diferentes valores de decisão de número de árvores, e 3 iterações com índices de vegetação e solo, para a produção do mapa LULC da região de Bragança (nordeste de Portugal). Finalmente, avaliámos a exactidão da classificação, antes e depois das tarefas de pós-classificação (generalização, fragmentação e remoção de pixels isolados).
Os resultados obtidos mostram que uma classificação com um nb-trees = 1000, incluindo índices de vegetação e solo, e após tarefas de pós-classificação, forneceu excelentes resultados de precisão (Coeficiente Kappa = 0.93, Precisão geral =96%, e erros marginais de omissão & comissão abaixo de 4%)
Random forest effectiveness for Bragança region mapping: comparing indices, number of the decision trees, and generalization
Get PDFMestrado de dupla diplomação com o Institute Agronomic and Veterinary Hassan IIRemote sensing is a domain that tends to use satellite images for classification and Land Use/Cover (LULC) mapping. For this purpose, classification algorithms are used, which are numerous and diverse, and it is necessary to establish decision criteria when choosing the algorithm. Ultimately, the main decision criterion will be the accuracy obtained in classification because the accuracy of classification may differ from one algorithm to another, even within the same algorithm, according to its variables. But there are other equally important criteria: it depends on the nature of the task, the quantity and types of data available, the type of response expected, the time and computational resources available, the depth of our knowledge about the algorithms.
The methodology of each part of the work was described and the criteria for comparison were established. In this research, with the same training data, the same validation data, the same application context (7 classes), and the same image data (Sentinel-2), we tested 15 iterations with the Random Forest classification algorithm, with different tree number decision values, and 3 iterations with vegetation and soil indexes, for the production of the LULC map of the Bragança region (northeast Portugal). Finally, we evaluate the accuracy of the classification, before and after the post-classification tasks (generalization, fragmentation and removal of isolated pixels).
The results obtained show that a classification with an nb-trees = 1000, including vegetation and soil indices, and after post-classification tasks, provided excellent precision results (Coefficient Kappa = 0.93, Overall accuracy = 96%, and marginal errors of omission & commission below 4%).A teledetecção é um domínio que tende a utilizar imagens de satélite para classificação e mapeamento de Uso/Cobertura da Terra (LULC). Para este fim, são utilizados algoritmos de classificação, que são numerosos e diversos, sendo necessário estabelecer critérios de decisão ao escolher o algoritmo. Em última análise, o principal critério de decisão será a precisão obtida na classificação, porque a precisão da classificação pode diferir de um algoritmo para outro, mesmo dentro do mesmo algoritmo, de acordo com as suas variáveis. Mas existem outros critérios igualmente importantes: depende da natureza da tarefa, da quantidade e tipos de dados disponíveis, do tipo de resposta esperada, do tempo e dos recursos computacionais disponíveis, da profundidade dos nossos conhecimentos sobre os algoritmos.
A metodologia de cada parte do trabalho foi descrita e os critérios de comparação foram estabelecidos. Nesta investigação, com os mesmos dados de formação, os mesmos dados de validação, o mesmo contexto de aplicação (7 classes), e os mesmos dados de imagem (Sentinel-2), testámos 15 iterações com o algoritmo de classificação Random Forest, com diferentes valores de decisão de número de árvores, e 3 iterações com índices de vegetação e solo, para a produção do mapa LULC da região de Bragança (nordeste de Portugal). Finalmente, avaliámos a exactidão da classificação, antes e depois das tarefas de pós-classificação (generalização, fragmentação e remoção de pixels isolados).
Os resultados obtidos mostram que uma classificação com um nb-trees = 1000, incluindo índices de vegetação e solo, e após tarefas de pós-classificação, forneceu excelentes resultados de precisão (Coeficiente Kappa = 0.93, Precisão geral =96%, e erros marginais de omissão & comissão abaixo de 4%)
Utilisation des Systèmes d’Informations Géographiques pour l’analyse spatio-temporelle des effets des changements climatiques sur la disponibilité des ressources halieutiques au Maroc
Get PDFChanges in climate variables such as temperature directly affect the habitats of marine species including fish species. In this study, we examinated the exposure level of distribution of three fish species, Sardina pilchardus, Merluccius merluccius and Thunnus thynnus. The study was carried out along Moroccan coasts in four main fishing areas. The exposure level to climate variables change was analysed based on variation of occurrence probabilties from reference period (2000-2014) and projected period (2040-2050). Two greenhouse gas emission scenarios were considered (RCP6.0 and RCP8.5). AquaMaps model was used to generate predicted distributions from three environmental factors, sea surface temperature, salinity and depth. The results demonstrate that temperature and salinity will increase during next 30 years in three areas (Mediterranean area, centre atlantic area and south atlantic area) under RCP6.0 and RCP8.5. Thus, distributions of Sardina pilchardus and Merluccius merluccius will decrease while the distribution of Thunnus thynnus will remain stable.
Keywords: Climate change, fisheries resources, GIS, AquaMaps, MoroccoLes changements des variables climatiques telles que la température affectent directement les conditions de vie des espèces marines y compris les espèces de poissons. Dans ce travail, nous avons examiné le niveau d’exposition de la distribution de trois espèces de Sardina pilchardus, Merluccius merluccius et Thunnus thynnus dans les quatre zones de pêche du Maroc. Le niveau d’exposition a été évalué par l’analyse des variations des probabilités d’occurrence des espèces sur une période de référence (2000-2014) et celle projetée (2040-2050) suivant les scénarios d’émissions des gaz à effet de serre (RCP6.0 et RCP 8.5). Le modèle de distribution utilisé dans ce travail est le modèle AquaMaps. Les facteurs écologiques considérés sont la température, la salinité et la profondeur. Les résultats ont montré que les variables environnementales (température et salinité) vont subir une augmentation au cours des trois prochaines décennies et selon les deux scénarios utilisés. La température et la salinité connaîtront une augmentation sur la majorité des zones, à l’exception de l’Atlantique Nord qui connaîtra une diminution de salinité. Par conséquent, la probabilité d’occurrence et la distribution des espèces de sardines et des merlus connaîtront une réduction tandis que celle du Thon rouge restera stable.
Mots clés : Changements climatiques, ressources halieutiques, SIG, AquaMaps, Maro
Going Beyond Counting First Authors in Author Co-citation Analysis
Get PDFThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Get PDF“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Get PDFWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Get PDFWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
- …
