This research examines how the harvesting behavior of nonindustrial private forest landowners, and their use of forestland for non-timber amenities, is affected by adjacent landowner behavior. The uncertainty an individual landowner has regarding adjacent landowners' preferences, and how the production of non-timber amenities on their own stands relies on the condition of adjacent stocks, is specifically addressed. Economic characterizations of substitutes and complements are employed to investigate the differences in optimal stock levels at the steady state in the production of amenities under various levels of cooperation among landowners. It is shown that there are externalities present when landowners do not coordinate management actions when parcels are spatially interdependent. The effects of spatial interdependencies on landowner behavior are further explored using data from a survey of forest landowners in Central Virginia. Findings suggest that forest landowners are willing to coordinate activities, and such decisions are determined by similar characteristics that function in predicting landowner behavior regarding timber harvesting. Further, landowners' decisions to use own and adjacent parcels were correlated, hinting at the spatial interdependencies of stocks in amenity valuations. Both the theoretical and empirical analyses suggest that the lack of coordination among landowners and its effects on stock management would be best addressed through the use of incentives to drive spatially efficient outcomes. / Ph. D.

nÃo hà / A anÃlise de imagens à uma tarefa fundamental em visÃo computacional. Ela influencia o desenvolvimento de algoritmos de processamento digital de imagens e as abordagens de avaliaÃÃo dos resultados produzidos por estes algoritmos. Esta tese introduz uma metodologia para a anÃlise estrutural de imagens baseada no uso combinado de uma transformaÃÃo estrutural multiescala e da extraÃÃo de caracterÃsticas estruturais por meio da matriz de interdependÃncia espacial. A transformaÃÃo estrutural multiescala à um algoritmo baseado no arcabouÃo da morfologia matemÃtica que mapeia os nÃveis de cinza da imagem de entrada para um espaÃo no qual esses nÃveis de cinza estÃo reagrupados em diferentes escalas de estruturas que formam os objetos. A transformaÃÃo pode ser aplicada no realce de imagens em nÃveis de cinza e na decomposiÃÃo de imagens binÃrias em estruturas elementares. A matriz de interdependÃncia espacial à um algoritmo baseado na estatÃstica de coocorrÃncia que produz uma representaÃÃo global das coincidÃncias das estruturas de duas imagens de entrada. Essa matriz provà quatro atributos, a saber, correlaÃÃo, momento de diferenÃa inverso, coeficiente chi-quadrado e entropia, os quais podem ser utilizados como descritores globais das estruturas da imagem. A metodologia proposta à validada com os resultados obtidos para diferentes aplicaÃÃes: a deteÃÃo de corrosÃo atmosfÃrica em fotografias de superfÃcies metÃlicas, a deteÃÃo de doenÃas pulmonares em imagens de tomografia computadorizada, a avaliaÃÃo referenciada da qualidade da imagens, a segmentaÃÃo dos vasos da retina em retinografias e a avaliaÃÃo da qualidade de algoritmos de segmentaÃÃo de vasos de retina. / Image analysis is a fundamental task in computer vision. It influences the development of algorithms for digital image processing and approaches for evaluating the results produced by these algorithms. This thesis introduces a methodology for the structural analysis of images based on the combined use of a multiscale structural transformation and extraction of structural features through spatial interdependence matrix. The multiscale structural transformation is an algorithm based on mathematical morphology framework that maps the gray levels of the input image into a space in which these gray levels are grouped into different scales of structures that form objects. The transformation can be applied in enhancement of gray level images and decomposition of binary images into elementary structures. The spatial interdependence matrix is an algorithm based on cooccurrence statistics that produces a global representation of the structural coincidences of two images. This matrix provides four attributes, namely, correlation, inverse difference moment, chi-square coefficient and entropy, which can be used as global descriptors of the image structures. The proposed methodology is validated with the results obtained for different applications: the detection of atmospheric corrosion of metal surfaces in photographs, the detection of lung disease in computerized tomography images, the referenced evaluation of image quality, the segmentation of retinal vessels in retinography and the quality assessment of retinal vessels segmentation algorithms.

Le changement structurel en agriculture suscite beaucoup d’intérêt de la part des économistes agricoles ainsi que des décideurs politiques. Pour prendre en compte l’hétérogénéité du comportement des agriculteurs, une approche par les modèles de mélange de chaînes de Markov est appliquée pour la première fois en économie agricole pour analyser ce processus. La performance de cette approche est d’abord testée en utilisant une forme simplifiée du modèle, puis sa forme générale est appliquée pour étudier l’impact de certaines mesures de politique agricole. Pour identifier les principaux canaux d’interdépendance entre exploitations voisines dans les processus du changement structurel, une approche de mélange non-Markovienne a été appliquée pour modéliser la survie et l’agrandissement des exploitations agricolesTrois principales conclusions découlent de cette thèse. Tout d’abord, la prise en compte de l’hétérogénéité dans les processus de transition des exploitations agricoles permet de mieux représenter le changement structurel et conduit à des prédictions plus précises de la distribution des exploitations, comparé aux modèles généralement utilisés jusqu’ici. Deuxièmement, l’impact des principaux facteurs du changement structurel dépend lui aussi des types non-observables d’exploitations mis en évidence. Enfin, le cadre du modèle de mélange permet également de révéler différents types de relations inobservées entre exploitations voisines qui contribuent au changement structurel observé à un niveau global ou régional. / Structural change in farming has long been the subject of considerable interest among agricultural economists and policy makers. To account for heterogeneity in farmers’ behaviours, a mixture Markov modelling framework is applied to analyse this process for the first time in agricultural economics. The performance of this approach is first investigated using a restrictive form of the model, and its general form is then applied to study the impact of some drivers of structural change, including agricultural policy measures. To identify channels through which interdependency between neighbouring farms arises in this process, the mixture modelling approach is applied to analyse both farm survival and farm growth. The main conclusions of this thesis are threefoldFirstly, accounting for the generally unobserved heterogeneity in the transition process of farms allows better representing structural change in farming and leads to more accurate predictions of farm-size distributions than the models usually used so far. Secondly, the impacts of the main drivers of structural change themselves depend on the specific unobservable farm types which are revealed by the model. Lastly, the mixture modelling approach enables identifying different unobserved relationships between neighbouring farms that contributes to the structural change observed at an aggregate or regional level.