Rehabilitating river corridors to restore valuable riparian habitat consumes significant resources from both governments and private companies. Given these considerable expenditures, it is important to monitor the progress of such projects. This study evaluated the utility of using Landsat Thematic Mapper remotely-sensed data from 2002 and 2009 to monitor vegetation change induced by instream flow restoration to the Lower Owens River in central California. This study compared the results of an unsupervised classification with an NDVI threshold classification to appraise the resources required and effectiveness of each analysis method. The results were inspected by creating standard remote sensing accuracy error matrices and by correlating landscape pattern metrics with bird indicator species. Both sets of classified maps show a noticeable increase in riparian vegetation in the study area following flow restoration in 2006, indicating an improvement of the quality of bird habitat. The study concluded that analyzing vegetation change using the unsupervised classification technique required more effort, expert knowledge, and supplementary data than using the NDVI threshold method. If these prerequisites are met, the output from the unsupervised classification process produces a more precise map of land cover change than the NDVI threshold method. However, if an analyst is lacking either resources or ground verification data, the NDVI threshold technique is capable of providing a generalized, but still valid evaluation of vegetation change. This conclusion is supported by higher correlations between indicator bird species under the unsupervised classification method than were found with the NDVI threshold method.

<p>  </p> <p>The domain of Deep Learning that is related to generation of textual description of images is called ‘Image Captioning.’ The central idea behind Image Captioning is to identify key features of an image and create meaningful sentences that describe the image. The current popular models include image captioning using Convolution Neural Network - Long Short-Term Memory (CNN-LSTM) based models and Attention based models. This research work first identifies the drawbacks of existing image captioning models namely – sequential style of execution, vanishing gradient problem and lack of context during training.</p> <p>This work aims at resolving the discovered problems by creating a Contextually Aware Image Captioning (CATIC) Model. The Transformer architecture, which solves the issues of vanishing gradients and sequential execution, forms the basis of the suggested model. In order to inject the contextualized embeddings of the caption sentences, this work uses Bidirectional Encoder Representation of Transformers (BERT). This work uses Remote Sensing Image Captioning Dataset. The results of the CATIC model are evaluated using BLEU, METEOR and ROGUE scores. On comparison the proposed model outperforms the CNN-LSTM model in all metrices. When compared to the Attention based model’s metrices, the CATIC model outperforms for BLEU2 and ROGUE metrices and gives competitive results for others.</p>

Fine-scale urban land cover information is important for a number of applications, including urban tree canopy mapping, green space analysis, and urban hydrologic modeling. Land cover information has traditionally been extracted from satellite or aerial images using automated image classification techniques, which classify pixels into different categories of land cover based on their spectral characteristics. However, in fine spatial resolution images (4 meters or better), the high degree of within-class spectral variability and between-class spectral similarity of many types of land cover leads to low classification accuracy when pixel-based, purely spectral classification techniques are used. Object-based classification methods, which involve segmenting an image into relatively homogeneous regions (i.e. image segments) prior to classification, have been shown to increase classification accuracy by incorporating the spectral (e.g. mean, standard deviation) and non-spectral (e.g. te xture, size, shape) information of image segments for classification. One difficulty with the object-based method, however, is that a segmentation parameter (or set of parameters), which determines the average size of segments (i.e. the segmentation scale), is difficult to choose. Some studies use one segmentation scale to segment and classify all types of land cover, while others use multiple scales due to the fact that different types of land cover typically vary in size. In this dissertation, two multi-scale object-based classification methods were developed and tested for classifying high resolution images of Deerfield Beach, FL and Houston, TX. These multi-scale methods achieved higher overall classification accuracies and Kappa coefficients than single-scale object-based classification methods. / Since the two dissertation methods used an automated algorithm (Random Forest) for image classification, they are also less subjective and easier to apply to other study areas than most existing multi-scale object-based methods that rely on expert knowledge (i.e. decision rules developed based on detailed visual inspection of image segments) for classifying each type of land cover. / by Brian A. Johnson. / Thesis (Ph.D.)--Florida Atlantic University, 2012. / Includes bibliography. / Electronic reproduction. Boca Raton, Fla., 2012. Mode of access: World Wide Web.

Image classification is a fundamental information extraction procedure in remote sensing that is used in land-cover and land-use mapping. Despite being considered as a replacement for manual mapping, it still requires some degree of analyst intervention. This makes the process of image classification time consuming, subjective, and error prone. For example, in unsupervised classification, pixels are automatically grouped into classes, but the user has to manually label the classes as one land-cover type or another. As a general rule, the larger the number of classes, the more difficult it is to assign meaningful class labels. A fully automated post-classification procedure for class labeling was developed in an attempt to alleviate this problem. It labels spectral classes by matching their spectral characteristics with reference spectra. A Landsat TM image of an agricultural area was used for performance assessment. The algorithm was used to label a 20- and 100-class image generated by the ISODATA classifier. The 20-class image was used to compare the technique with the traditional manual labeling of classes, and the 100-class image was used to compare it with the Spectral Angle Mapper and Maximum Likelihood classifiers. The proposed technique produced a map that had an overall accuracy of 51%, outperforming the manual labeling (40% to 45% accuracy, depending on the analyst performing the labeling) and the Spectral Angle Mapper classifier (39%), but underperformed compared to the Maximum Likelihood technique (53% to 63%). The newly developed class-labeling algorithm provided better results for alfalfa, beans, corn, grass and sugar beet, whereas canola, corn, fallow, flax, potato, and wheat were identified with similar or lower accuracy, depending on the classifier it was compared with. / vii, 93 leaves : ill., maps (some col.) ; 29 cm

Thesis (Ph.D.) -- Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Thesis (Ph.D.)--Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Thesis (MSc)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: Supervised classifiers are commonly used in remote sensing to extract land cover information. They are, however, limited in their ability to cost-effectively produce sufficiently accurate land cover maps. Various factors affect the accuracy of supervised classifiers. Notably, the number of available training samples is known to significantly influence classifier performance and to obtain a sufficient number of samples is not always practical. The support vector machine (SVM) does perform well with a limited number of training samples. But little research has been done to evaluate SVM’s performance for geographical object-based image analysis (GEOBIA). GEOBIA also allows the easy integration of additional features into the classification process, a factor which may significantly influence classification accuracies. As such, two experiments were developed and implemented in this research. The first compared the performances of object-based SVM, maximum likelihood (ML) and nearest neighbour (NN) classifiers using varying training set sizes. The effect of feature dimensionality on classifier accuracy was investigated in the second experiment. A SPOT 5 subscene and a four-class classification scheme were used. For the first experiment, training set sizes ranging from 4-20 per land cover class were tested. The performance of all the classifiers improved significantly as the training set size was increased. The ML classifier performed poorly when few (<10 per class) training samples were used and the NN classifier performed poorly compared to SVM throughout the experiment. SVM was the superior classifier for all training set sizes although ML achieved competitive results for sets of 12 or more training samples per class. Training sets were kept constant (20 and 10 samples per class) for the second experiment while an increasing number of features (1 to 22) were included. SVM consistently produced superior classification results. SVM and NN were not significantly (negatively) affected by an increase in feature dimensionality, but ML’s ability to perform under conditions of large feature dimensionalities and few training areas was limited. Further investigations using a variety of imagery types, classification schemes and additional features; finding optimal combinations of training set size and number of features; and determining the effect of specific features should prove valuable in developing more costeffective ways to process large volumes of satellite imagery. KEYWORDS Supervised classification, land cover, support vector machine, nearest neighbour classification maximum likelihood classification, geographic object-based image analysis / AFRIKAANSE OPSOMMING: Gerigte klassifiseerders word gereeld aangewend in afstandswaarneming om inligting oor landdekking te onttrek. Sulke klassifiseerders het egter beperkte vermoëns om akkurate landdekkingskaarte koste-effektief te produseer. Verskeie faktore het ʼn uitwerking op die akkuraatheid van gerigte klassifiseerders. Dit is veral bekend dat die getal beskikbare opleidingseenhede ʼn beduidende invloed op klassifiseerderakkuraatheid het en dit is nie altyd prakties om voldoende getalle te bekom nie. Die steunvektormasjien (SVM) werk goed met beperkte getalle opleidingseenhede. Min navorsing is egter gedoen om SVM se verrigting vir geografiese objek-gebaseerde beeldanalise (GEOBIA) te evalueer. GEOBIA vergemaklik die integrasie van addisionele kenmerke in die klassifikasie proses, ʼn faktor wat klassifikasie akkuraathede aansienlik kan beïnvloed. Twee eksperimente is gevolglik ontwikkel en geïmplementeer in hierdie navorsing. Die eerste eksperiment het objekgebaseerde SVM, maksimum waarskynlikheids- (ML) en naaste naburige (NN) klassifiseerders se verrigtings met verskillende groottes van opleidingstelle vergelyk. Die effek van kenmerkdimensionaliteit is in die tweede eksperiment ondersoek. ʼn SPOT 5 subbeeld en ʼn vier-klas klassifikasieskema is aangewend. Opleidingstelgroottes van 4-20 per landdekkingsklas is in die eerste eksperiment getoets. Die verrigting van die klassifiseerders het beduidend met ʼn toename in die grootte van die opleidingstelle verbeter. ML het swak presteer wanneer min (<10 per klas) opleidingseenhede gebruik is en NN het, in vergelyking met SVM, deurgaans swak presteer. SVM het die beste presteer vir alle groottes van opleidingstelle alhoewel ML kompeterend was vir stelle van 12 of meer opleidingseenhede per klas. Die grootte van die opleidingstelle is konstant gehou (20 en 10 eenhede per klas) in die tweede eksperiment waarin ʼn toenemende getal kenmerke (1 tot 22) toegevoeg is. SVM het deurgaans beter klassifikasieresultate gelewer. SVM en NN was nie beduidend (negatief) beïnvloed deur ʼn toename in kenmerkdimensionaliteit nie, maar ML se vermoë om te presteer onder toestande van groot kenmerkdimensionaliteite en min opleidingsareas was beperk. Verdere ondersoeke met ʼn verskeidenheid beelde, klassifikasie skemas en addisionele kenmerke; die vind van optimale kombinasies van opleidingstelgrootte en getal kenmerke; en die bepaling van die effek van spesifieke kenmerke sal waardevol wees in die ontwikkelling van meer koste effektiewe metodes om groot volumes satellietbeelde te prosesseer. TREFWOORDE Gerigte klassifikasie, landdekking, steunvektormasjien, naaste naburige klassifikasie, maksimum waarskynlikheidsklassifikasie, geografiese objekgebaseerde beeldanalise

Models of land use change fall into two broad categories: pattern based and process based. This thesis focuses on pattern based land use change models, expanding our understanding of these models in three important ways. First, it is demonstrated that some driving variables do not have a smooth impact on the land use transition process. Our example variable is access to water. Land managers with access to piped water do not have any need for surface or groundwater. For variables like this, a model needs to change the way that driving variables are represented. The second important result is that including a variable which captures spatial correlation between land use types significantly increases the explanatory power of the prediction model. A major weakness of pattern based land use models is their inability to model interactions between neighbouring land parcels; the method proposed in this study can be an alternative to account for spatial neighbourhood association. These innovations are applied using the CLUE-S (Conversion of Land Use and its Effects at Small regional extent) system to the Deep Creek watershed in the Southern Interior of British Columbia. The results highlight the challenge of balancing the protection of agricultural land and conserving forest and natural areas when population and economic growth are inevitable. The results also demonstrate the implications of land use change on existing land use policies. The calibrated model was validated using remote sensing data. A series of discriminant functions were estimated for each land use type in the recent period and these functions were then used to classify. The calibrated model was run in reverse, back to the generated land use classification, and results compared. Fit was reasonable with error rates falling below ten percent when radii beyond 2.5 km were considered. Another important contribution is demonstrating the importance of modelling dynamic variables. Some important drivers are changing continuously and others depend on land use change itself. Failure to update these variables will bias model forecasts. Spatial neighbourhood association, an endogenous variable governed by land use change itself, is again used as the example dynamic variable. The study demonstrates the importance of updating all associated information. / Graduate Studies, College of (Okanagan) / Graduate

Orientadores: Alexandre Xavier Falcão, Jefersson Alex dos Santos / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-27T14:43:51Z (GMT). No. of bitstreams: 1 VargasMunoz_JohnEdgar_M.pdf: 9138091 bytes, checksum: bdb40e3a5655df0e10a137f2d08f0d8d (MD5) Previous issue date: 2015 / Resumo: Recentemente, técnicas de aprendizado de máquina têm sido propostas para criar mapas temáticos a partir de imagens de sensoriamento remoto. Estas técnicas podem ser divididas em métodos de classificação baseados em pixels ou regiões. Este trabalho concentra-se na segunda abordagem, uma vez que estamos interessados em imagens com milhões de pixels e a segmentação da imagem em regiões (superpixels) pode reduzir consideravelmente o número de amostras a serem classificadas. Porém, mesmo utilizando superpixels, o número de amostras ainda é grande para anotá-las manualmente e treinar o classificador. As técnicas de aprendizado ativo propostas resolvem este problema começando pela seleção de um conjunto pequeno de amostras selecionadas aleatoriamente. Tais amostras são anotadas manualmente e utilizadas para treinar a primeira instância do classificador. Em cada iteração do ciclo de aprendizagem, o classificador atribui rótulos e seleciona as amostras mais informativas para a correção/confirmação pelo usuário, aumentando o tamanho do conjunto de treinamento. A instância do classificador é melhorada no final de cada iteração pelo seu treinamento e utilizada na iteração seguinte até que o usuário esteja satisfeito com o classificador. Observamos que a maior parte dos métodos reclassificam o conjunto inteiro de dados em cada iteração do ciclo de aprendizagem, tornando este processo inviável para interação com o usuário. Portanto, enderaçamos dois problemas importantes em classificação baseada em regiões de imagens de sensoriamento remoto: (a) a descrição efetiva de superpixels e (b) a redução do tempo requerido para seleção de amostras em aprendizado ativo. Primeiro, propusemos um descritor contextual de superpixels baseado na técnica de sacola de palavras, que melhora o resultado de descritores de cor e textura amplamente utilizados. Posteriormente, propusemos um método supervisionado de redução do conjunto de dados que é baseado em um método do estado da arte em aprendizado ativo chamado Multi-Class Level Uncertainty (MCLU). Nosso método mostrou-se tão eficaz quanto o MCLU e ao mesmo tempo consideravelmente mais eficiente. Adicionalmente, melhoramos seu desempenho por meio da aplicação de um processo de relaxação no mapa de classificação, utilizando Campos Aleatórios de Markov / Abstract: In recent years, machine learning techniques have been proposed to create classification maps from remote sensing images. These techniques can be divided into pixel- and region-based image classification methods. This work concentrates on the second approach, since we are interested in images with millions of pixels and the segmentation of the image into regions (superpixels) can considerably reduce the number of samples for classification. However, even using superpixels the number of samples is still large for manual annotation of samples to train the classifier. Active learning techniques have been proposed to address the problem by starting from a small set of randomly selected samples, which are manually labeled and used to train a first instance of the classifier. At each learning iteration, the classifier assigns labels and selects the most informative samples for user correction/confirmation, increasing the size of the training set. An improved instance of the classifier is created by training, after each iteration, and used in the next iteration until the user is satisfied with the classifier. We observed that most methods reclassify the entire pool of unlabeled samples at every learning iteration, making the process unfeasible for user interaction. Therefore, we address two important problems in region-based classification of remote sensing images: (a) the effective superpixel description and (b) the reduction of the time required for sample selection in active learning. First, we propose a contextual superpixel descriptor, based on bag of visual words, that outperforms widely used color and texture descriptors. Second, we propose a supervised method for dataset reduction that is based on a state-of-art active learning technique, called Multi-Class Level Uncertainty (MCLU). Our method has shown to be as effective as MCLU, while being considerably more efficient. Additionally, we further improve its performance by applying a relaxation process on the classification map by using Markov Random Fields / Mestrado / Ciência da Computação / Mestre em Ciência da Computação

Orientador: Maria de Lourdes Bueno Trindade Galo / Resumo: O represamento de rios para a produção de energia elétrica usualmente provoca atividades antrópicas que impactam um ecossistema aquático fortemente. Uma das consequências de se instalar pequenos reservatórios em regiões sujeitas à intensos processos de urbanização e industrialização é a abundância de macrófitas, resultante do despejo de nutrientes em grandes concentrações no ecossistema aquático. Recentemente, o grande volume de images multitemporais de sensoriamento remoto disponíveis em bancos de dados gratuitos, bem como a alta performance computacional que permite a mineração de grandes volumes de dados, fazem com que o monitoramento de fenômenos ambientais seja um objeto de estudo recorrente. O propósito desse estudo é desenvolver uma metodologia baseada na integração de dados heterogêneos, fornecidos por séries temporais de coleções de imagens multiespectrais e multitemporais Landsat e coleções de dados climáticos históricos, para investigar a evolução e comportamento espacial de macrófitas aquáticas em lagos e reservatórios eutrofizados. A extensa coleção temporal de imagens de superfície de reflectância Landsat disponível e também dados de variáveis ambientais permitiram a construção e análise de séries temporais para investigar a recorrente abundância de macrófitas no reservatório de Salto Grande, localizado na região metropolitana de Campinas, São Paulo, Brasil. Inicialmente, foi encontrado que as imagens Landsat possuem a qualidade radiométrica necessária para se r... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: River damming for electric power production usually triggers anthropic activities that strongly impact on aquatic ecosystem. One of the consequences of installing small reservoirs in regions subject to an intense process of urbanization and industrialization is the overabundance of macrophytes, resulting from the input of nutrients in high concentration into the aquatic ecosystem. Currently, the large volume of multitemporal remote sensing images available in open data sources, as well as the high computational performance that allow the mining of large volumes of data has made the monitoring of environmental phenomena a recurrent object of analysis. The aim of this study is to develop a methodology based on the integration of heterogeneous data, provided by time series of multispectral and multitemporal Landsat images and collections of historical climatic data, to investigate the evolution and spatial behavior of aquatic macrophytes in lakes and eutrophic reservoirs. So, the extensive temporal collection of the Landsat surface reflectance images made available as well as environmental variables data permitted the construction and analysis of time series to investigate the recurrent over-abundance of macrophytes in Salto Grande reservoir, located in the metropolitan region of Campinas, São Paulo, Brazil. Initially, it was found that the the Landsat images have the necessary radiometric quality to perform the time series analyses, through an assessment based on information ab... (Complete abstract click electronic access below) / Doutor