Search

This study focused on the structural differences of lignin after pyridine&ndash, acetic acid&ndash, water (PAW) and dioxane&ndash, acidic water (DAW) purification processes. These structural differences included the S/G ratio, condensed structure, weight-average (MW) molecular weights, &beta, O-4 linkages and sugar content. The chemical structure of the isolated crude lignin (CL), PAW purified lignin (PPL) and DAW purified lignin (DPL) was elucidated using quantitative 13C NMR, 2D-HSQC NMR spectra, thermogravimetric analysis (TGA), gel permeation chromatography (GPC) and Fourier transform infrared spectroscopy (FTIR). The results showed that the PPL fractions contain fewer condensed structures, higher S/G ratios, more &beta, O-4 linkages, higher average MW and lower thermal degradation properties compared to the CL and DPL fractions. Furthermore, the PAW process was more selective in removing condensed units and enriching S-type lignin from CL compared to the DAW process. These results provide valuable information for understanding which purification process is more suitable to be applied for lignin.

Protein Chemistry Laboratory, Department of Chemistry, Bose Institute, 9311, Acharya Prafulla Chandra Road, Kolkata-700 009, India Manuscrtpt received 17 October 2011, accepted 20 October 2011 ��-Crystallin is the major constituent protein of the eye lens of the vertebrates. It is an oligomenc protein having micelle like architecture. Ever since it was reported in early 90's that it possesses molecular chaperone like function which is crucial for the maintenance of the transparency of the eye lens, interests in understanding the mechanism of such function developed rapidly. Various laboratories have contributed towards understanding of the structure and function of this protein. Our group has been engaged to study specific problems related to this area. We undertook detatled study to understand the structure of bound substrates employing a combination of simple molecular biology and bio physical tools to show that ��-crystallin recognizes early folding intermediates and bound substrates native-like structure. Our work showed that chaperone activity correlated fairly well with exposed surface hydrophobicity. We proved that temperature activation was not required for ��-crystallin to function as chaperone as was proposed by others. Very importantly we demonstrated that some small molecules such as ATP can interact with a-crystallin to form weak associa ted chaperone-substrate complex that can play crucial role in increasing the thermodynamic stability and chaperone function and refractive properties of the lens. Some bivalent metal ions such as zn2+ plays a very important role in the structure, stability and chaperone function of a-crystallin and such a sore has considerable physiological sigmficance in understanding the cataract formation in the lens. This article is not intended to be a comprehensive account of all aspects of work done on this protein. But instead in this article we have reviewed primarily our work and mentioned the work done by others in these specific areas only.

International audience; Collections of documents are sets of heterogeneous documents, like a specific ancient book series, having proper structural and semantic properties linking them. A particular collection contains document images with specific physical layouts, like text pages or full-page illustrations, appearing in a specific order. Its contents, like journal articles, may be shared by several pages, not necessary following, producing strong dependencies between pages interpretations.In order to build an analysis system which can bring contextual information from the collection to the appropriate recognition modules for each page, we propose to express the structural and the semantic properties of a collection with a definite clause grammar. This is made possible by representing collections as streams of document descriptors, and by using extensions to the formalism we present here. We are then able to automatically generate a parser dedicated to a collection. Beside allowing structural variations and complex information flows, we also show that this approach enables the design of analysis stages, on a document or a set of documents. The interest of context usage is illustrated with several examples and their appropriate formalization in this framework.

http://www.ieee.org; In this paper we present an innovative approach to automatically generate adjacency grammars describing graphical symbols. A grammar production is formulated in terms of rulesets of geometrical constraints among symbol primitives. Given a set of symbol instances sketched by a user using a digital pen, our approach infers the grammar productions consisting of the ruleset most likely to occur. The performance of our work is evaluated using a comprehensive benchmarking database of on-line symbols.

Our work has been motivated by the issue of generic handwritten characters recognition. We try to address it with structural methods based on graph modelling. The documents processed are unconstrained and come from different periods. Classical statistical methods are efficients but they can only process languages with restrained vocabulary according to a learning phase. Two recognition systems based on attributed graphs are proposed. The first one uses numerical attributes and random graphs for modelling the learning base. The structural information changes the complexity notion and allows an interesting cooperation with statistical methods. The second one uses hierarchical fuzzy attributes. It is a model-based recognition system with no learning phase. It brings an interesting first step for generic recognition.; L'objectif des travaux réalisés au cours de cette thèse est d'adresser la problématique de la reconnaissance générique de caractères manuscrits par les méthodes structurelles à base de graphes. Les écrits traités sont non-contraints et hétérogènes dans le temps. Les méthodes classiques, dites statistiques, sont efficaces mais ne peuvent s'appliquer qu'à des écritures à vocabulaire restreint dans le cadre d'un système avec une phase d'apprentissage. Nous proposons deux systèmes de reconnaissance à base de graphes d'attributs. Le premier utilise des attributs numériques et une modélisation de la base d'apprentissage avec des graphes aléatoires. L'intégration des informations de structure change la notion de complexité et permet une coopération intéressante avec les approches statistiques. Le second système utilise des attributs hiérarchiques flous. Il permet une reconnaissance sans apprentissage basée sur des modèles qui tend vers la reconnaissance générique recherchée.

Our work has been motivated by the issue of generic handwritten characters recognition. We try to address it with structural methods based on graph modelling. The documents processed are unconstrained and come from different periods. Classical statistical methods are efficients but they can only process languages with restrained vocabulary according to a learning phase. Two recognition systems based on attributed graphs are proposed. The first one uses numerical attributes and random graphs for modelling the learning base. The structural information changes the complexity notion and allows an interesting cooperation with statistical methods. The second one uses hierarchical fuzzy attributes. It is a model-based recognition system with no learning phase. It brings an interesting first step for generic recognition.; L'objectif des travaux réalisés au cours de cette thèse est d'adresser la problématique de la reconnaissance générique de caractères manuscrits par les méthodes structurelles à base de graphes. Les écrits traités sont non-contraints et hétérogènes dans le temps. Les méthodes classiques, dites statistiques, sont efficaces mais ne peuvent s'appliquer qu'à des écritures à vocabulaire restreint dans le cadre d'un système avec une phase d'apprentissage. Nous proposons deux systèmes de reconnaissance à base de graphes d'attributs. Le premier utilise des attributs numériques et une modélisation de la base d'apprentissage avec des graphes aléatoires. L'intégration des informations de structure change la notion de complexité et permet une coopération intéressante avec les approches statistiques. Le second système utilise des attributs hiérarchiques flous. Il permet une reconnaissance sans apprentissage basée sur des modèles qui tend vers la reconnaissance générique recherchée.

This thesis describes the study and implementation of a component for structural recognition of handwritten or typesetted mathematical formulae. This work is related to document analysis and recognition fields of research. Our aim could be resumed as : how to recognize the mathematical notation, i.e. the abstract syntax tree of a formula, just with graphical and geometrical informations (symbols and their position). The design of our software, OFR (Optical Formula Recognition) enables us to reuse the same prototype with different OCR systems and to adapt the recognition process to a large variety of mathematical notations. To achive the structural recognition, we used a graph modelisation. This allows us to have all data in a synthetized format. We use an attributed contextual graph grammar for parsing graph, especially developped for mathematical operators. Thanks to formulae exchange protocols like OpenMath, the interface developped with the help of OFR may be used as a front end for writing mathematical ans send them to a computer algebra system like Mathematica.; Le sujet de ce mémoire est l'étude et la réalisation d'un composant pour la reconnaissance structurelle des formules mathématiques typographiées et manuscrites. Ces travaux s'inscrivent dans une thématique plus large : l'analyse et la reconnaissance de documents. La problématique générale que nous avons considérée peut se résumer de la manière suivante ; il s'agit d'identifier la structure, ou arbre de syntaxe abstraite, d'une formule à partir des données graphiques et géométriques (les symboles composant la notation et leur position). L'architecture logicielle retenue permet d'adapter très facilement le composant, baptisé OFR (Reconnaissance Optique de Formules), aux logiciels fournissant les symboles, ainsi qu'aux diverses notations mathématiques identifiées. Pour effectuer cette reconnaissance structurelle, nous avons eu recours à une modélisation à base de graphes. Elle permet une abstraction des données receuillies et une transformation de ces informations par la définition d'une grammaire de graphes contextuelle attribuée, spécialement adaptée aux opérateurs mathématiques. En nous appuyant sur des protocoles de communication d'objets mathématiques, comme OpenMath, nous pouvons envisager l'utilisation de l'interface développée autour d'OFR comme une alternative à la saisie des formules mathématiques.

In einer Literaturübersicht wird der aktuelle Stand bei der Entwicklung wissensbasierter Werkstückerkennungssysteme dargestellt, der dadurch gekennzeichnet ist, daß es an Verfahren zur automatischen Modellerstellung mangelt. Es wird deshalb ein wissensbasiertes Werkstückerkennungssystem beschrieben, das Rückkopplungsschleifen für die automatische Modellerstellung enthält. Bei der automatischen Modellerstellung werden in einem iterativen Optimierungsvorgang Modellbeschreibungselemente ausgewählt, Filterparameter an die zu erkennenden Werkstücke angepaßt und eine Erkennungsstrategie ermittelt, die angibt, wie der Vergleichsvorgang bei der Erkennung eines Werkstückes ablaufen soll. Das System wurde für die Erstellung von 2 D-Modellen entwickelt; die Erweiterbarkeit auf 3 D-Modelle ist im Systementwurf berücksichtigt und wurde durch Konstruktion einfacher 3 D-Modelle bereits nachgewiesen. Mit den automatisch erstellten und optimierten Modellen können teilweise verdeckte Werkstücke erkannt werden, eine Aufgabe, die mit konventionellen Mustererkennungsverfahren ohne Modellsteuerung nicht gelöst werden kann.