This thesis investigates the use of combined field and satellite data for crop identification and area estimation in County Durham, Northeast England. The satellite data were obtained by the Thematic Mapper (TM) sensor onboard Landsat-5 on 31 May 1985. The TM data were geometrically corrected to the British National Grid and the county boundaries were digitized in order to apply the methodology used in this study on a county basis. The field data were obtained by applying a stratified random sampling strategy. The area was subdivided into five main strata and forty four 1km(_^2) sample units were randomly chosen and fully surveyed by the author using a pre-prepared questionnaire. The field area measurements were taken and the final hectarage estimates were obtained for each crop. The research demonstrated the ability of Landsat-TM data to discriminate between agricultural crops in the study area. Results obtained emphasised that satellite data can be used for identification of agricultural crops over large geographic areas with small field sizes and different environmental and physical features. A land-cover classification system appropriate to the study area was designed. Using the Landsat-TM data, the study produced a classification map of thirteen land-cover types with more than 80% accuracy. The classification accuracy was assessed quantitatively by using the known land-use information obtained from the sample units visited during the field survey. The study analysed the factors which influenced the degree of separability between different agricultural crops since some crops were more clearly identified than others. Using a double sampling method based on the combination of both Landsat- TM and field data in regression analysis, a hectarage estimate was produced for each crop type in County Durham. The results obtained showed that the regression estimator was always more efficient than the field estimator. Crop area estimated by regression reduced the imprecision in all strata and was more efficient in some strata than others. This indicated that a gain in precision was achieved by using Landsat- TM in conjunction with the field data. The results illustrated that stratification based on an environmental criterion was an efficient approach as far as the the application of agricultural remote sensing in County Durham is concerned. The stratified approach allowed each stratum to be analysed separately, thereby lessening the reliance on cloud free imagery for the whole county on any given date. Furthermore, the results obtained by this study suggest that it is possibile to link remote sensing data with existing county based information systems on agricultural and land-use.

Cette recherche s’inscrit dans le cadre de l’élaboration d’espaces numériques d’interaction tactile, en tant qu’ils sont rendus possibles par la connexion en réseau d’un dispositif de suppléance perceptive. Dans ce contexte, nous avons articulé une recherche appliquée, centrée sur les analyses d’usage du dispositif dans des contextes écologiques, à une recherche fondamentale, dirigée par les questions théoriques soulevées par les usages eux-mêmes.Pour cela, la méthodologie minimaliste nous offre une occasion originale d’étudier la constitution même des rencontres interpersonnelles, à travers un espace de couplage sensorimoteur inédit. Nos résultats généraux suggèrent que le processus de l’interaction est une dynamique relationnelle autonome qui émerge de l’engagement mutuel des participants. Dans une approche énactive et interactionniste de la cognition sociale, il s’agit alors de considérer que le croisement perceptif entre deux sujets, même réduit à son expression la plus simple, consiste en la rencontre de deux activités perceptives construisant du sens. Dans le contexte de rencontres strictement dyadiques, nous tentons de caractériser cette dynamique, de façon à en isoler les éléments constitutifs. Nous montrons alors que la coordination interpersonnelle résulte d’un processus actif de co-ajustements dynamiques qui se joue autant à un niveau microscopique (qualité de l’accroche perceptive) qu’à un niveau macroscopique (organisation de séquences d’interaction). Dans le contexte de l’exploration mutuelle d’un contenu numérique, nous montrons que cette dynamique d’interaction permet aux participants de faire sens de leurs engagements respectifs en fonction des objets présents. La coordination des activités perceptives se présente alors comme le support de la co-constitution d’un monde partagé de significations, à travers une compréhension interpersonnelle ancrée dans un contexte pragmatique. En outre, l’articulation des expérimentations et des analyses d’usage écologiques nous a conduits à proposer des spécifications techniques et fonctionnelles pour les espaces numériques partagés, de façon à proposer un dispositif pertinent pour les utilisateurs déficients visuels. / Our research is part of the design of tactile interaction digital spaces, as they are made possible by a network connection between perceptual supplementation devices. Within this framework, we articulated an applied research, focused on the analysis of use of the device in ecological contexts, with a fundamental one, led by theoretical issues raised by the usage itself. In this respect, minimalist methodology gives us an opportunity to study the very constitution of interpersonal encounters, through a new space of sensorimotor coupling for users.Our overall results suggest that interaction process is an autonomous relational dynamics that emerges from the mutual engagement of participants. Within an enactive and interactionist approach of social cognition, we consider that this perceptual crossing, even when reduced to its simplest expression, involves the meeting of two perceptual activities which inherently produce sense. In the context of strictly dyadic encounters, we try to characterize this dynamic, so as to isolate its fundamental features. We show that interpersonal coordination results from an active process of dynamic co-adjustments that unfolds both on a microscopic level (quality of the perceptual coupling) and on a macroscopic level (organization of interaction sequences). In the case of mutual exploration of digital content, we show that this dynamic interaction allows participants to make sense of their respective engagement relatively to the present objects. The coordination of perceptual activities is then presented as a support for the co-constitution of a shared world of meanings, from an interpersonal understanding rooted in a pragmatic context.In addition, these experiments, coupled to ecological usage analysis, have led us to propose technical and functional specifications for shared digital spaces, so as to provide a relevant device for visually impaired users.

Dans cette thèse, nous traitons de la compréhension de contenus visuels, qu’il s’agisse d’images, de vidéos ou encore de contenus 3D. On entend par compréhension la capacité à inférer des informations sémantiques sur le contenu visuel. L’objectif de ce travail est d’étudier des méthodes combinant deux approches : 1) l’analyse automatique des contenus et 2) l’analyse des interactions liées à l’utilisation de ces contenus (analyse des usages, en plus bref). Dans un premier temps, nous étudions l’état de l’art issu des communautés de la vision par ordinateur et du multimédia. Il y a 20 ans, l’approche dominante visait une compréhension complètement automatique des images. Cette approche laisse aujourd’hui plus de place à différentes formes d’interventions humaines. Ces dernières peuvent se traduire par la constitution d’une base d’apprentissage annotée, par la résolution interactive de problèmes (par exemple de détection ou de segmentation) ou encore par la collecte d’informations implicites issues des usages du contenu. Il existe des liens riches et complexes entre supervision humaine d’algorithmes automatiques et adaptation des contributions humaines via la mise en œuvre d’algorithmes automatiques. Ces liens sont à l’origine de questions de recherche modernes : comment motiver des intervenants humains ? Comment concevoir des scénarii interactifs pour lesquels les interactions contribuent à comprendre le contenu manipulé ? Comment vérifier la qualité des traces collectées ? Comment agréger les données d’usage ? Comment fusionner les données d’usage avec celles, plus classiques, issues d’une analyse automatique ? Notre revue de la littérature aborde ces questions et permet de positionner les contributions de cette thèse. Celles-ci s’articulent en deux grandes parties. La première partie de nos travaux revisite la détection de régions importantes ou saillantes au travers de retours implicites d’utilisateurs qui visualisent ou acquièrent des con- tenus visuels. En 2D d’abord, plusieurs interfaces de vidéos interactives (en particulier la vidéo zoomable) sont conçues pour coordonner des analyses basées sur le contenu avec celles basées sur l’usage. On généralise ces résultats en 3D avec l’introduction d’un nouveau détecteur de régions saillantes déduit de la capture simultanée de vidéos de la même performance artistique publique (spectacles de danse, de chant etc.) par de nombreux utilisateurs. La seconde contribution de notre travail vise une compréhension sémantique d’images fixes. Nous exploitons les données récoltées à travers un jeu, Ask’nSeek, que nous avons créé. Les interactions élémentaires (comme les clics) et les données textuelles saisies par les joueurs sont, comme précédemment, rapprochées d’analyses automatiques des images. Nous montrons en particulier l’intérêt d’interactions révélatrices des relations spatiales entre différents objets détectables dans une même scène. Après la détection des objets d’intérêt dans une scène, nous abordons aussi le problème, plus ambitieux, de la segmentation. / This thesis focuses on the problem of understanding visual contents, which can be images, videos or 3D contents. Understanding means that we aim at inferring semantic information about the visual content. The goal of our work is to study methods that combine two types of approaches: 1) automatic content analysis and 2) an analysis of how humans interact with the content (in other words, usage analysis). We start by reviewing the state of the art from both Computer Vision and Multimedia communities. Twenty years ago, the main approach was aiming at a fully automatic understanding of images. This approach today gives way to different forms of human intervention, whether it is through the constitution of annotated datasets, or by solving problems interactively (e.g. detection or segmentation), or by the implicit collection of information gathered from content usages. These different types of human intervention are at the heart of modern research questions: how to motivate human contributors? How to design interactive scenarii that will generate interactions that contribute to content understanding? How to check or ensure the quality of human contributions? How to aggregate human contributions? How to fuse inputs obtained from usage analysis with traditional outputs from content analysis? Our literature review addresses these questions and allows us to position the contributions of this thesis. In our first set of contributions we revisit the detection of important (or salient) regions through implicit feedback from users that either consume or produce visual contents. In 2D, we develop several interfaces of interactive video (e.g. zoomable video) in order to coordinate content analysis and usage analysis. We also generalize these results to 3D by introducing a new detector of salient regions that builds upon simultaneous video recordings of the same public artistic performance (dance show, chant, etc.) by multiple users. The second contribution of our work aims at a semantic understanding of fixed images. With this goal in mind, we use data gathered through a game, Ask’nSeek, that we created. Elementary interactions (such as clicks) together with textual input data from players are, as before, mixed with automatic analysis of images. In particular, we show the usefulness of interactions that help revealing spatial relations between different objects in a scene. After studying the problem of detecting objects on a scene, we also adress the more ambitious problem of segmentation.

The development of shape in multicellular organisms has intrigued human minds for millenia. Empowered by modern genetic techniques, molecular biologists are now striving to not only dissect developmental processes, but to exploit their modularity for the design of custom living systems used in bioproduction, remediation, and regenerative medicine. Currently, our capacity to harness this potential is fundamentally limited by a lack of spatiotemporal control over gene expression in multicellular systems. While several synthetic genetic circuits for control of multicellular patterning have been reported, hierarchical induction of gene expression domains has received little attention from synthetic biologists, despite its fundamental role in biological self-organization. In this thesis, I introduce the first synthetic genetic system implementing population-based AND logic for programmed hierarchical patterning of bacterial populations of Escherichia coli, and address fundamental prerequisites for implementation of an analogous genetic circuit into the emergent multicellular plant model Marchantia polymorpha. In both model systems, I explore the use of bacteriophage T7 RNA polymerase as a gene expression engine to control synthetic patterning across populations of cells. In E. coli, I developed a ratiometric assay of bacteriophage T7 RNA polymerase activity, which I used to systematically characterize different intact and split enzyme variants. I utilized the best-performing variant to build a three-color patterning system responsive to two different homoserine lactones. I validated the AND gate-like behavior of this system both in cell suspension and in surface culture. Then, I used the synthetic circuit in a membrane-based spatial assay to demonstrate programmed hierarchical patterning of gene expression across bacterial populations. To prepare the adaption of bacteriophage T7 RNA polymerase-driven synthetic patterning from the prokaryote E. coli to the eukaryote M. polymorpha, I developed a toolbox of genetic elements for spatial gene expression control in the liverwort: I analyzed codon usage across the transcriptome of M. polymorpha, and used insights gained to design codon-optimized fluorescent reporters successfully expressed from its nuclear and chloroplast genomes. For targeting of bacteriophage T7 RNA polymerase to these cellular compartments, I functionally validated nuclear localization signals and chloroplast transit peptides. For spatiotemporal control of bacteriophage T7 RNA polymerase in M. polymorpha, I characterized spatially restricted and inducible promoters. For facilitated posttranscriptional processing of target transcripts, I functionally validated viral enhancer sequences in M. polymorpha. Taking advantage of this genetic toolbox, I introduced inducible nuclear-targeted bacteriophage T7 RNA polymerase into M. polymorpha. I showed implementation of the bacteriophage T7 RNA polymerase/PT7 expression system accompanied by hypermethylation of its target nuclear transgene. My observations suggest operation of efficient epigenetic gene silencing in M. polymorpha, and guide future efforts in chassis engineering of this multicellular plant model. Furthermore, my results encourage utilization of spatiotemporally controlled bacteriophage T7 RNA polymerase as a targeted silencing system for functional genomic studies and morphogenetic engineering in the liverwort. Taken together, the work presented enhances our capacity for spatiotemporal gene expression control in bacterial populations and plants, facilitating future efforts in synthetic morphogenesis for applications in synthetic biology and metabolic engineering.