No description available.

No description available.

Foreground extraction technology plays an important role in image and video processing tasks. It has been widely used in various industries. To better describe the overlap relationship between foreground and background, alpha channel is introduced. It reveals the opacity property of foreground objects. Thus, fully extracting a foreground object requires determining the alpha values for pixels, also known as extracting an alpha matte. In this thesis, we propose an improved sampling-based alpha matting algorithm, which is capable of generating high quality matting results. By analyzing the weakness of previous approaches, we optimize the sampling process and consider the cost of each sample pair to avoid missing any good samples. The good performance is demonstrated even for complex images. On the other hand, extracting foreground objects from video sequences is a more challenging task since it has higher demands on accuracy and efficiency. Previous approaches usually require a significant amount of user input and the results still suffer from inaccuracy. In this thesis, we successfully extend our algorithm to video sequences and let it run in an automatic fashion. Adaptive trimap, which is vital for matting, can be automatically generated and properly propagated in this system. Our method not only reduces the user interference but also guarantees the matting quality.

No description available.

We have observed two high -redshift quasars with the echelle spectrograph and 2D- Frutti Photon Counter at Las Campanas, in order to investigate the statistical properties of the Lyman -a forest. The two-point correlation function for the Lyman -a forest lines at z .^s 3 is consistent with zero, for all velocity splittings A > 50 km /sec. When Lyman -a lines and other metal lines from known metalline systems are included, the correlation function shows a weak non -zero signal at small A . We suggest that the weak clustering of the Lyman -a forest detected by other workers may be the result of contamination by a small number of metal -line systems and their associated Lyman -a lines.

No description available.

The chloride intracellular channel (CLIC) family of proteins are an unusual class of chloride channels that possess the ability to auto-insert into cellular membranes. The CLICs exhibit ubiquitous tissue and cellular distributions and adopt a glutathione S-transferase fold in the soluble form that is highly conserved in vertebrates. CLIC homologues have been identified in the model organisms C. elegans and D. melanogaster, and in the former case have been extensively characterised in regards to function. In this thesis, we present the crystal structure of the D. melanogaster CLIC, revealing several unique features in the conserved invertebrate CLIC fold including an elongated C-terminal extension and metal binding site. The bound metal is identified as the potassium cation, resolving concerns regarding previously published work that assign the metal as the isoelectronic calcium cation. It has been reported that a human CLIC protein, CLIC4, translocates to the nucleus in response to cellular stress, facilitated by a putative CLIC4 nuclear localisation signal (NLS). The CLIC4 NLS adopts α-helical structure in the native CLIC4 structural fold. It is proposed that CLIC4 is transported to the nucleus via the classical nuclear import pathway after binding the import receptor, importin-α. We have determined the X-ray crystal structure of a truncated form of importin-α bound to a CLIC4 NLS bearing peptide. The NLS peptide binds the major binding site in an extended conformation similar to that observed for the classical SV40 large T-antigen NLS. A tyrosine residue within the CLIC4 NLS makes surprisingly favourable interactions by forming side chain hydrogen bonds to the importin-α backbone. This structural evidence supports the hypothesis that CLIC4 translocation to the nucleus is governed by the importin-α nuclear import pathway, providing it can undergo a conformational rearrangement that exposes the NLS in an extended conformation. We further analyse importin-α:NLS binding interactions by solving high resolution structures of truncated importin-α containing an empty binding site and bound to the SV40 NLS. A surprising interaction is discovered between importin-α and an NLS-like motif in the endogenous E. coli 30S ribosomal subunit S21, revealing new insight into importin-α recognition of full length cargo.

La triade non-catalytique est un motif strictement conservé des alpha-amylases chlorure-dépendentes qui est parfaitement superposable avec la triade catalytique des protéases à sérine active. Le but de ce travail était de déterminer le rôle de cette triade. Par des expériences de mutagenèse, nous avons pu montrer que ce rôle est de nature structurale. Des expériences de RMN nous ont permis de démontrer la présence d'un pont H anormalement fort dans ces enzymes, ce qui pourrait expliquer l'instabilité très marquée des mutants de la triade. Malheureusement, nous n'avons pas pu attribuer sans ambiguité ce pont H à la triade non-catalytique. La dernière partie de ce travail décrit la recherche de la triade non-catalytique dans des protéines autres que les amylases.

Létude des protéines amyloïdogéniques représente un intérêt fondamental car ces protéines subissent une transconformation et une agrégation qui sont étroitement liées à lapparition de maladies incurables telles que la maladie dAlzheimer, de Parkinson ou de Creutzfeldt-Jakob. Ces phénomènes ne sont pas encore complètement expliqués tant au niveau énergétique que structural. Les peptides obliques sont de courts fragments protéiques de 11 à 19 acides aminés qui sont capables de sinsérer obliquement dans les membranes biologiques et de les déstabiliser. Récemment, des peptides obliques ont été mis en évidence dans deux protéines amyloïdogéniques responsables de maladies neurodégénératives. Il sagit du peptide b amyloïde, qui cause la maladie dAlzheimer, et de la protéine PrP, agent de la maladie de Creutzfeldt-Jakob. Au sein des protéines amyloïdogéniques, les peptides obliques pourraient être impliqués dans leffet neurotoxique de ces protéines. En affectant directement la membrane cellulaire grâce à leurs propriétés déstabilisatrices, ils conduiraient à la mort cellulaire. Dautre part, les peptides obliques pourraient aussi être impliqués dans le processus de transconformation de ces protéines. Le but de ce travail est de mettre en évidence des fragments obliques parmi dautres protéines amyloïdogéniques décrites dans la littérature, par des méthodes de modélisation moléculaire puis détudier expérimentalement les propriétés de plusieurs de ces peptides vis-à-vis des liposomes, ainsi que leur structure et leur toxicité. Ce travail a également permis de poser les bases dune méthode de détection automatique des peptides obliques. A lissue de cette étude, 22 peptides obliques, appartenant à 18 protéines amyloïdogéniques différentes, ont pu être mis en évidence parmi un ensemble de 53 protéines étudiées, et 7 peptides obliques ont été étudiés expérimentalement. Les résultats renforcent lhypothèse dune implication de certains peptides obliques dans les phénomènes de transconformation et/ou de toxicité associés aux protéines amyloïdogéniques.

Foreground extraction technology plays an important role in image and video processing tasks. It has been widely used in various industries. To better describe the overlap relationship between foreground and background, alpha channel is introduced. It reveals the opacity property of foreground objects. Thus, fully extracting a foreground object requires determining the alpha values for pixels, also known as extracting an alpha matte. In this thesis, we propose an improved sampling-based alpha matting algorithm, which is capable of generating high quality matting results. By analyzing the weakness of previous approaches, we optimize the sampling process and consider the cost of each sample pair to avoid missing any good samples. The good performance is demonstrated even for complex images. On the other hand, extracting foreground objects from video sequences is a more challenging task since it has higher demands on accuracy and efficiency. Previous approaches usually require a significant amount of user input and the results still suffer from inaccuracy. In this thesis, we successfully extend our algorithm to video sequences and let it run in an automatic fashion. Adaptive trimap, which is vital for matting, can be automatically generated and properly propagated in this system. Our method not only reduces the user interference but also guarantees the matting quality.