Thèse numérisée par la Division de la gestion de documents et des archives de l'Université de Montréal

Single view reconstruction is fundamentally an under-constrained problem. We aim to develop new approaches to model human face and motion with model priors that restrict the space of possible solutions. First, we develop a novel approach to recover the 3D shape from a single view image under challenging conditions, such as large variations in illumination and pose. The problem is addressed by employing the techniques of non-linear manifold embedding and alignment. Specifically, the local image models for each patch of facial images and the local surface models for each patch of 3D shape are learned using a non-linear dimensionality reduction technique, and the correspondences between these local models are then learned by a manifold alignment method. Local models successfully remove the dependency of large training databases for human face modeling. By combining the local shapes, the global shape of a face can be reconstructed directly from a single linear system of equations via least square. Unfortunately, this learning-based approach cannot be successfully applied to the problem of human motion modeling due to the internal and external variations in single view video-based marker-less motion capture. Therefore, we introduce a new model-based approach for capturing human motion using a stream of depth images from a single depth sensor. While a depth sensor provides metric 3D information, using a single sensor, instead of a camera array, results in a view-dependent and incomplete measurement of object motion. We develop a novel two-stage template fitting algorithm that is invariant to subject size and view-point variations, and robust to occlusions. Starting from a known pose, our algorithm first estimates a body configuration through temporal registration, which is used to search the template motion database for a best match. The best match body configuration as well as its corresponding surface mesh model are deformed to fit the input depth map, filling in the part that is occluded from the input and compensating for differences in pose and body-size between the input image and the template. Our approach does not require any makers, user-interaction, or appearance-based tracking. Experiments show that our approaches can achieve good modeling results for human face and motion, and are capable of dealing with variety of challenges in single view reconstruction, e.g., occlusion.

At certain stages of a bumblebee colony life cycle workers lay eggs. Not all workers reproduce, however, since many continue to forage and care for the nest. This leads to questions regarding what differentiates a reproductive worker from a non-reproductive one. It is hypothesized that a form of reproductive competition takes place, where the most behaviourally dominant worker becomes reproductively dominant. The behaviour of orphaned Bombus impatiens pairs was recorded and aggression, social interactions, egg-laying, and ovarian development were identified. Experiment 1 examined the association between aggression and egg-laying. Contrary to the hypothesis, the most aggressive worker did not lay more eggs. When the ovarian development of workers was manipulated and two workers with developed ovaries were paired (Experiment 3), they were more aggressive than pairs with discouraged ovarian development. This provides support for the supposition that aggression and reproduction are related, however, it is only partial support as worker pairs with encouraged ovarian development did not lay more eggs. Since aggression is believed to be only one part of behavioural dominance, Experiment 2 studied the association between social interactions and aggression and reproduction. Results showed that when two socially active bees were paired they were more aggressive than pairs including one or two socially inactive bumblebees. No significant difference in ovarian development between socially active pairs and socially inactive pairs was found. Brood presence was also predicted to affect reproductive control. Experiment 1 found egg-laying and aggression were more likely to co-occur in the absence of brood. Results from Experiment 2 supplemented the first experiment since the absence of brood increased rates of aggression and ovarian development in pairs. Whereas the results confirm aggression has a role in worker reproduction the findings also reveal that behavioural dominance does not equate to reproductive dominance under all conditions. The primary contributions of this thesis were the development of a method to distinguish behavioural dominance from reproductive dominance and determining their relationship under different environments (brood presence) and experimental manipulations (ovarian development). These contributions further define dominance in Bombus impatiens.

In this era of increasing standard of living and rapid growth of civil engineering construction, environmental issues pertaining to natural resources depletion, global warming, energy crisis, waste pollution and greenhouse gas emission have been major issues of concern throughout the world and most especially in the construction industry. This research was conducted to investigate the applicability of recycled wastepaper as lightweight building materials with focus on contributing to sustainability in the creation of the built environment. The major aim was to develop an eco-friendly lightweight non-loadbearing block from recycled wastepaper without the use of cement as binder. This study specifically addressed the drawback of low strength development that usually occur with increasing wastepaper content in the existing cement-based-wastepaper blocks. It also indirectly addresses; the environmental impacts associated with the construction industry (including; high consumption of natural resources, greenhouse gas emission, high energy consumption and so on), the environmental pollution resulting from unsustainable waste generation, and the generic drawback of high water absorption that plagues wastepaper-based blocks. To achieve this, research methods including; laboratory experimentation and simulation modelling were employed. The research outcome is an eco-friendly block unit designated as Cement-less Wastepaper-based Lightweight Block (CWLB) which contains 75% waste content and exhibiting properties that satisfy the requirements for application as non-loadbearing lightweight blocks in building construction. CWLB displayed compressive strength that far outweighs those recorded for the existing cement-based wastepaper blocks available in the literature. The properties recorded for the optimal CWLB includes; 2.71 MPa average compressive strength, 901.5 kg/m3 average density, 0.19 W/m.k thermal conductivity, 989.9 m/s ultrasonic pulse velocity, 0.0026 g/m2.S0.5 average coefficient of capillary water absorption and 883.38 MPa estimated elastic modulus. The approximate compressive strength of 2.38 MPa and 1.58 MPa were respectively predicted and recorded for the solid and hollow finite element model samples of CWLB. The impressive satisfactory properties of CWLB for the intended application and its eco-friendliness in terms of natural resources conservation and improved compressive strength suggests that CWLB shall indeed serve as a more sustainable alternative to the reigning/existing cement-based-wastepaper blocks and to the conventional masonry blocks of the same category. Amongst other things, future work will address the validation of the approximate compressive strength predicted for the solid and hollow CWLB insitu samples in order to take further the subject matter.

Målet med denna uppsats är att bidra med förklarande kunskap om hur en transition från traditionella systemutvecklingsmetoder till agila systemutvecklingsmetoder kan genomföras, vilka problem som kan uppstå under övergången samt de utmaningar organisationen ställs inför vid en övergång. Vidare är målet också att bidra med vägledande kunskap i den mån att den ska belysa möjliga fallgropar som kan undvikas och vilka aspekter som är viktiga att ta hänsyn till vid en övergång till en agil metod. Detta genomförs genom en förklarande fallstudie (Oates, p. 143, 2006) av övergången till agil systemutveckling från en annan, icke-agil metod. Den slutgiltiga uppsatsen ämnar att bidra med en kunskapsgrund till i vilka förhållanden det finns en möjlighet att byta till en agil metod. Undersökningens resultat visar att det det finns goda möjligheter för en övergång men att det är viktigt att både de som driver förändringen och övriga involverade har förståelse för vad som kommer att förändras. Framförallt människors ovilja att lämna bekväma strukturer och ta sig an nytt ansvar är problem som en organisation behöver vara medveten om och har förståelse för ifall en övergång ska lyckas. På en större skala är det också extra viktigt för organisationen att besitta de rätta förutsättningarna för en övergång utöver en djupare kunskap i de olika variationerna av skalning som tas upp i uppsatsen.

Esta pesquisa trata do projeto de redes de distribuição de carga parcelada. Mais especificamente são tratados dois tipos de problemas que são comuns no planejamento desse tipo de sistema. O primeiro deles corresponde ao problema estratégico de configuração de redes do tipo hub-and-spoke, consistindo na definição simultânea da quantidade e localização de terminais para consolidação de carga (ou hubs), e na definição da alocação dos terminais aos hubs localizados. Uma vez determinada a configuração da rede, o segundo problema, no nível de decisão tático, corresponde na definição do caminho que cada carga parcelada deve percorrer desde sua origem até alcançar seu terminal de destino, a um mínimo custo, tendo a rede hub-and-spoke como um dado de entrada do problema. Um novo modelo matemático é proposto para representar o problema estratégico de configuração de uma rede hub-and-spoke, possuindo uma menor quantidade de variáveis e restrições, ao se comparar com outros modelos matemáticos comumente utilizados para representar o problema. Esse novo modelo matemático permitiu a obtenção de soluções ótimas para problemas em redes com até 100 terminais, sendo apresentada pela primeira vez a solução ótima para problemas utilizados como benchmark na literatura. Dado que problemas de grande porte ainda continuam muito difíceis de serem resolvidos, são propostas três variantes de uma heurística simples e eficiente utilizando técnicas de multi-início e busca tabu, bem como uma heurística integrada em dois estágios baseada em busca tabu para solução. Experimentos computacionais utilizando dados tradicionalmente utilizados na literatura para solução de problemas de configuração de redes hub-and-spoke (conjuntos de dados CAB e AP), bem como instâncias novas e modificadas, mostraram que a abordagem utilizada para solução do problema possibilitou a obtenção da solução ótima, ou a melhor solução conhecida, para esses problemas em um tempo de processamento muito curto, permitindo assim resolver de forma eficiente problemas de grande porte, nunca antes resolvidos em pesquisas anteriores. O segundo problema foi motivado por uma aplicação prática de uma empresa de transporte rodoviário de cargas parceladas no Brasil. O problema diz respeito ao planejamento de carregamentos a serem realizados em cada terminal, levando-se em consideração cada carga parcelada que precisa ser transportada, definindo o percurso que cada carga deve percorrer até chegar ao seu destino. É proposto um modelo matemático e, dada a dificuldade para se resolver problemas de tamanho como o encontrado na prática, é proposto também um método de solução utilizando metaheurística busca tabu. Experimentos computacionais realizados mostraram que a heurística proposta pôde efetivamente resolver problemas de tamanho como o encontrado na prática. / This research deals with problems related to distribution networks for less-than-truckload (LTL) freight transportation. More specifically, we deal with two relevant problems that arise. The first corresponds to the strategic problem of designing and configuring hub-and-spoke networks in terms of simultaneously determining the optimal number of consolidation terminals (hub) nodes, their locations and the allocation of the other terminals (spokes) to the hubs. . Once the network configuration is determined, the second problem, in the tactical level of decision, corresponds to defining the path that each LTL individual freight needs to follow from its origin to reach its destination terminal, at a minimum cost, having a hub-and-spoke network topology as a data entry to the problem. A new mathematical model is proposed to represent the strategic problem of designing a hub-and-spoke network, with fewer variables and constraints than previous formulations found in the literature This model allowed us to obtain optimal solutions for problems in transportation networks with up to 100 terminals, reporting for the first time the optimal solutions of benchmark problems in the literature. Since this problems still remains too hard to solve for larger instances, we propose we propose three variants of a simple and efficient multi-start tabu search heuristic as well as a two-stage integrated tabu search heuristic to solve it. Computational experiments using typical benchmark problems (CAB and AP data sets) as well as new and modified instances show that our approaches consistently return the optimal or best-known results in very short CPU times, thus allowing the possibility of efficiently solving larger instances of the USAHLP than those found in the literature. The second problem is motivated by a practical application of a LTL transportation company in Brazil. It deals with the planning of loads to be done at each terminal, taking into account each LTL freight that needs to be transported, defining the path that each good needs to follow to reach its destination. A new mathematical model is proposed, and, since real world problems are very hard to solve, a heuristic based on tabu search is also developed. Computational experiments show that our heuristic can effectively solve real-world instances from a trucking company in Brazil.

Os fitoplasmas são organismos procariotos sem parede celular, que habitam as células do floema das plantas e são transmitidos, principalmente, por cigarrinhas sugadoras de floema. Este patógeno é responsável por causar doenças em diversas espécies vegetais, provocando a expressão de diversos sintomas, entre os quais podem ser destacados a redução do tamanho foliar, amarelecimento das folhas, superbrotamento de ramos e enfezamento da planta hospedeira. Entre os anos de 2013 e 2014, plantas de três espécies vegetais com sintomas característicos de infecção causada por fitoplasmas foram observadas no estado de São Paulo. O DNA total foi extraído a partir de plantas de couve-flor com sintomas de nanismo, malformação da inflorescência e necrose dos vasos do floema; de plantas da ornamental conhecida por árvore-da-felicidade com sintomas amarelecimento e redução do limbo foliar; e de plantas de Alho-poró com sintomas de enfezamento. Por meio do teste de PCR conduzido em sua maioria com os primers P1A/16S-SR foi possível detectar fitoplasmas nas três espécies vegetais testadas. Os fragmentos genômicos correspondentes ao 16S rRNA dos fitoplasmas foram sequenciados e identificados. Nas plantas de couve-flor e da árvore-da-felicidade foram identificados fitoplasmas afilados ao grupo 16SrVII-B, através da análise virtual de RFLP, cálculo do coeficientes de similaridade (F) e análise filogenética. Para ambas as espécies, este é o primeiro relato da ocorrência de representantes deste grupo, nas condições brasileiras. Nas plantas de Alho-poró foram identificados fitoplasmas afiliados ao subgrupo 16SrIII-J e ao grupo 16SrXIII, com base na análise dos genes 16S rRNA, SecY e proteína ribossomal, conduzidas através de análise virtual de RFLP, valores de coeficientes de similaridade e análise filogenética. Para o fitoplasma membro do grupo 16SrXIII foram identificadas quatro estirpes, uma delas afiliada ao subgrupo 16SrXIII-E e as demais como prováveis representantes de novos subgrupos. O presente estudo se constitui em uma contribuição ao conhecimento de novos patossistemas envolvendo fitoplasmas, bem como à caracterização molecular de fitoplasmas baseadas em diferentes genes marcadores, atualmente usados para a classificação de representantes deste grupo emergente de agentes causais de doenças de plantas. / The phytoplasmas are prokaryotic organisms without cell walls, which inhabit the phloem cells of plants and are transmitted mainly by leafhoppers sucking the phloem. This pathogen is responsible for causing diseases in various plant species, leading to expression of many symptoms, among which can be highlighted the reduction of leaf size, leaf yellowing, shoots proliferation and stunting of the plant host. Between the years 2013 and 2014, three plant species with characteristic symptoms of infection caused by phytoplasma were observed in São Paulo. Total DNA was extracted from cauliflower plants with symptoms of stunting, malformation of the inflorescence and necrosis of the phloem; ornamental plants known for Ming Aralia with symptoms yellowing and little leaves; and leek plants with symptoms of stunting. Through the PCR test conducted mostly with the primers P1A/16S-SR was possible to detect phytoplasmas in the three species of plants. The genomic fragments corresponding of 16S rRNA gene of the phytoplasma were sequenced and identified. In plants of cauliflower and Ming Aralia were identified phytoplasmas belonging to 16SrVII-B group through virtual RFLP analysis, calculation of similarity coefficients (F) and phylogenetic analysis. For both species, this is the first report of the occurrence of representatives of this group, in Brazilian conditions. In leek plants were identified phytoplasmas affiliated with 16SrIII-J subgroup, and the 16SrXIII group, based on the analysis of the 16S rRNA, SecY and ribosomal protein genes, conducted through virtual analysis of RFLP, similarity coefficient values and phylogenetic analysis. For the phytoplasma member of group 16SrXIII were identified four strains, one affiliated with 16SrXIII-E subgroup and the others as probable representatives of new subgroups. This study constitutes a contribution to knowledge of new pathosystems involving phytoplasmas and the molecular characterization of phytoplasma based on different marker genes, currently used for the classification of representatives of this emerging group of plant pathogens.

Cette thèse de création porte sur les nouveaux enjeux de pesanteur etde légèreté revendiqués par un certain nombre d’artistes contemporains.Nous vivons dans un monde de plus en plus pesant et pourtant les artistesexplorent des moyens de s’extraire de cette gravité pour atteindreune apesanteur théorique. Car l’apesanteur telle que nous nous la représentonsn’existe pas en terme de physique : il s’agit d’un phénomène dechute libre qui crée l’impression de flottement. J’interroge donc la notiond’(a)pesanteur, qui prend en considération cette dialectique de l’envol etde la chute, au sein de nombreuses oeuvres, principalement sculpturales. Jequestionne ma pratique de la sculpture, de l’installation et du dessin à traversle prisme de l’(a)pesanteur, qui construit un véritable rapport aumonde et à l’acte de création. J’utilise les propriétés intrinsèques du matériauafin de générer des sensations de poids ou de légèreté. J’interrogeégalement l’acte de suspendre une oeuvre dans l’espace d’exposition: lasuspension serait-elle aujourd’hui devenue le nouveau socle de la sculpture? De l’espace d’exposition à l’Espace cosmique, l’oeuvre d’art est, depuisla conquête spatiale, de plus en plus présente hors des frontières terrestres.Elle est créée dans l’Espace ou pour l’Espace : l’Espace cosmique serait-il devenu le nouveau lieu de l’art ?A travers les figures de la chute,de l’envol,du suspens,de l’imaginaire spatial,et jusqu’aux figures de l’invisible, l’(a)pesanteur tisse un véritable filrouge à travers la création contemporaine. / This creative thesis is about the new challenges of weightlessness andlightness faced by a number of contemporary artists.We are living in a increasinglyheavy world, yet artists are exploring ways to escape this gravityto reach a theoretical weightlessness. Because the way to display weightlessnessdoesn’t exist in terms of physics: it’s the phenomenon of free-fallingwhich creates this feeling of floating. So I seek to test theweight(less)ness concept, which explores this float and fall dialectic in numerousworks, mainly sculptural. I’m questioning my own practice ofsculpture, installation and drawing, through the prism of weight(less)ness,which builds an actual connection to the world and to the act of creation.I’m using the intrinsic property of the material to generate feelings ofweight or lightness. I’m also questioning the act of hanging a work in theexhibition space : nowadays, has suspension become the new pedestal of asculpture ? From the exhibition space to cosmic Space, artwork increasinglygains presence beyond earth’s boundaries, ever since the conquest ofspace. It’s created in Space and for Space : has cosmic Space become thenew place for art? Through figures of falling, of flight, of suspension, ofimaginary spaces, and even figures of invisibility, weight(less)ness builds averitable guiding light through contemporary creation.

The aim of this dissertation is to develop a new algorithm to embed a watermark in JPEG compressed images, using encoding methods. This encompasses the embedding of proprietary information, such as identity and authentication bitstrings, into the compressed material. This watermark encoding scheme involves combining entropy coding with homophonic coding, in order to embed a watermark in a JPEG image. Arithmetic coding was used as the entropy encoder for this scheme. It is often desired to obtain a robust digital watermarking method that does not distort the digital image, even if this implies that the image is slightly expanded in size before final compression. In this dissertation an algorithm that combines homophonic and arithmetic coding for JPEG images was developed and implemented in software. A detailed analysis of this algorithm is given and the compression (in number of bits) obtained when using the newly developed algorithm (homophonic and arithmetic coding). This research shows that homophonic coding can be used to embed a watermark in a JPEG image by using the watermark information for the selection of the homophones. The proposed algorithm can thus be viewed as a ‘key-less’ encryption technique, where an external bitstring is used as a ‘key’ and is embedded intrinsically into the message stream. The algorithm has achieved to create JPEG images with minimal distortion, with Peak Signal to Noise Ratios (PSNR) of above 35dB. The resulting increase in the entropy of the file is within the expected 2 bits per symbol. This research endeavor consequently provides a unique watermarking technique for images compressed using the JPEG standard. / Dissertation (MEng)--University of Pretoria, 2008. / Electrical, Electronic and Computer Engineering / unrestricted

At certain stages of a bumblebee colony life cycle workers lay eggs. Not all workers reproduce, however, since many continue to forage and care for the nest. This leads to questions regarding what differentiates a reproductive worker from a non-reproductive one. It is hypothesized that a form of reproductive competition takes place, where the most behaviourally dominant worker becomes reproductively dominant. The behaviour of orphaned Bombus impatiens pairs was recorded and aggression, social interactions, egg-laying, and ovarian development were identified. Experiment 1 examined the association between aggression and egg-laying. Contrary to the hypothesis, the most aggressive worker did not lay more eggs. When the ovarian development of workers was manipulated and two workers with developed ovaries were paired (Experiment 3), they were more aggressive than pairs with discouraged ovarian development. This provides support for the supposition that aggression and reproduction are related, however, it is only partial support as worker pairs with encouraged ovarian development did not lay more eggs. Since aggression is believed to be only one part of behavioural dominance, Experiment 2 studied the association between social interactions and aggression and reproduction. Results showed that when two socially active bees were paired they were more aggressive than pairs including one or two socially inactive bumblebees. No significant difference in ovarian development between socially active pairs and socially inactive pairs was found. Brood presence was also predicted to affect reproductive control. Experiment 1 found egg-laying and aggression were more likely to co-occur in the absence of brood. Results from Experiment 2 supplemented the first experiment since the absence of brood increased rates of aggression and ovarian development in pairs. Whereas the results confirm aggression has a role in worker reproduction the findings also reveal that behavioural dominance does not equate to reproductive dominance under all conditions. The primary contributions of this thesis were the development of a method to distinguish behavioural dominance from reproductive dominance and determining their relationship under different environments (brood presence) and experimental manipulations (ovarian development). These contributions further define dominance in Bombus impatiens.