While uncertainty in scientific data attracts an increasing research interest in the visualization community, two critical issues remain insufficiently studied: (1) visualizing the impact of the uncertainty of a data set on its features and (2) interactively exploring 3D or large 2D data sets with uncertainties. In this study, a suite of feature-based techniques is developed to address these issues. First, a framework of feature-level uncertainty visualization is presented to study the uncertainty of the features in scalar and vector data. The uncertainty in the number and locations of features such as sinks or sources of vector fields are referred to as feature-level uncertainty while the uncertainty in the numerical values of the data is referred to as data-level uncertainty. The features of different ensemble members are indentified and correlated. The feature-level uncertainties are expressed as the transitions between corresponding features through new elliptical glyphs. Second, an interactive visualization tool for exploring scalar data with data-level and two types of feature-level uncertainties — contour-level and topology-level uncertainties — is developed. To avoid visual cluttering and occlusion, the uncertainty information is attached to a contour tree instead of being integrated with the visualization of the data. An efficient contour tree-based interface is designed to reduce users’ workload in viewing and analyzing complicated data with uncertainties and to facilitate a quick and accurate selection of prominent contours. This thesis advances the current uncertainty studies with an in-depth investigation of the feature-level uncertainties and an exploration of topology tools for effective and interactive uncertainty visualizations. With quantified representation and interactive capability, feature-based visualization helps people gain new insights into the uncertainties of their data, especially the uncertainties of extracted features which otherwise would remain unknown with the visualization of only data-level uncertainties.

<p>Two important deficiencies have been identified for the integration of CAD and automated process planning. These deficiencies stem from the lack of a uniform representation scheme of pans and products, and an effective communication for CAD and process planning. This thesis presents a new approach and original knowledge regarding the integration and individual aspects of feature-based design, cellular manufacturing planning, inspection planning and assembly sequence planning.</p> <p>A high-level new language called Feature-based Design Description Language (FDDL) has been proposed and designed with a feature representation scheme. Its syntax, semantics and vocabulary have been defined with consideration given to the user, the engineering terminology, and the computer implementation. The FDDL system consists of a number of lexical analyzers, a parser and three code generators. Once the products or parts modeled by the FDDL, or by a feature-based modeler, are processed using the FDDL system, inputs are created for manufacturing tasks planning systems.</p> <p>A feature-based modeling and manufacturing tasks planning system has been designed and implemented, and consists of a prototype of a feature-based modeler, the FDDL system, a feature-based cellular manufacturing planning system, a feature-based automated inspection task planner, and a prototype assembly sequence planner. The prototype feature-based modeler is used to model components using features. All expert tolerancing consultant module has been included in the modeler to assist the user. Cellular manufacturing planning deals with group formation and parts assignment to cells. A clustering-based optimization approach has been proposed and implemented for the formation of machine cells and part families. A feature-based assignment system has been developed to integrate the feature-based design and the formed cells. Automata and pattern recognition techniques, in combination with manufacturing knowledge, are used in the system. The feature-based inspection planner has been developed to integrate the feature-based design and a Coordinate Measuring Machine (CMM). Original inspection strategies and knowledge have been developed for CMM, based on the analysis of CMM characteristics, tolerancing theories, features representation, part structure and geometry. A knowledge-based approach has been presented to integrate CAD with the assembly sequence planning. A prototype of such an assembly sequence planner has been developed for generating the assembly sequence for products from the design directly.</p> / Doctor of Philosophy (PhD)

A demanda por mais flexibilidade nas fábricas e serviços originou um aumento no volume de operações internas de carga e descarga, devido à maior diversidade dos elementos transportados. Logo, na busca por um fluxo de materiais mais eficiente, as empresas passaram a investir em soluções tecnológicas, entre elas, o uso de Automated Guided Vehicles (AGVs), por conta do custo mais atrativo e do avanço em relação aos primeiros AGVs, que até então dependiam de uma infraestrutura adicional para suportar a navegação. Muitos AGVs modernos possuem movimentação livre e são orientados por sistemas que utilizam sensores para interpretar o ambiente, sendo assim, tornar os AGVs autônomos despertou o interesse de pesquisadores na área de robótica móvel para o desenvolvimento de sistemas capazes de auxiliar e coordenar a navegação. Novas técnicas de localização, tal como a localização baseada em marcadores reflexivos, e a construção de armazéns com layouts estruturados para a navegação viabilizaram o uso de AGVs autônomos, entretanto sua utilização em armazéns existentes ainda é um desafio. Neste contexto, o Laboratório de Robótica Móvel (LabRom) do Grupo de Mecatrônica da EESC/USP, através do projeto do Armazém Inteligente, tem pesquisado os problemas de: roteamento, gerenciamento das baterias, navegação e auto-localização. Robôs autônomos precisam de um sistema de auto-localização eficiente e preciso para navegar com segurança, o qual depende de um mapa e da interpretação do ambiente utilizando sensores embarcados. Para alcançar esse objetivo este trabalho propõe um Sistema de Auto-localização baseado no Extended Kalman Filter (EKF) como solução. O sistema, desenvolvido em linguagem C, interage com outros dois sistemas: roteamento e navegação e foi implementado em um armazém simulado utilizando o software Player/Stage, mostrando ser confiável no fornecimento de uma estimativa de localização baseada em odometria e landmarks com localização conhecida. O sistema foi novamente testado utilizando a odometria de um robô móvel Pioneer P3-AT e os valores de um sensor de medição laser 2D SICK LMS200 extraídos de um ambiente indoor real. Para este teste foi construído um feature-based map a partir de um desenho de planta baixa no formato CAD e utilizou-se o algoritmo de segmentação Iterative End-Point Fit (IEPF) para interpretar o ambiente. Os resultados mostraram que as vantagens oferecidas pelas características padronizadas de um ambiente indoor, semelhante a um armazém, podem viabilizar o uso do Sistema de Auto-localização em armazéns existentes. / The demand for more flexibility in factories and services led to an increase in the volume of internal operations of loading and unloading, due to the greater diversity of elements transported. Hence, in the search for a more efficient materials flow, companies went to invest in technology solutions, among them, the use of Automated Guided Vehicles (AGVs), on account of the more attractive cost and improvement over the first AGVs, which hitherto depended of an additional infrastructure to support navigation. Many modern AGVs have free movement and are guided by systems that use sensors to interpret the environment, thus make AGVs autonomous aroused the interest of researchers in the mobile robotics field to development of systems able to assist and coordinate the navigation. New localization techniques, such as localization based on reflective markers, and the construction of warehouses with structured layouts for navigation did feasible the use of autonomous AGVs, however its use in existing warehouses is still a challenge. In this context, the Mobile Robotics Lab (LabRom) of the Mechatronics Group of EESC/USP, through the Intelligent Warehouse Project, has researched the problems: routing, battery management, navigation and self-localization. Autonomous robots need an efficient and accurate self-localization system to safely navigate, which depends on one map and of the interpretation of the environment using embedded sensors. To achieve this goal, this work proposes a Self-Localization System based on the Extended Kalman Filter (EKF) as a solution. The system, developed in C language, interacts with two other systems: routing and navigation and was implemented in a simulated warehouse using the Player/Stage software, showing to be reliable in providing an estimative of localization based on odometry and landmarks with known localization. The system was again tested using the odometry of mobile robot Pioneer P3-AT and the values of a 2D Laser Rangefinder SICK LMS200 extracted from a real indoor environment. For this test was built a feature-based map from a floor plan design in CAD format and was used the segmentation algorithm Iterative End-Point Fit (IEPF) to interpret the environment. The results showed that the advantages offered by the standard features of indoor environment, like a warehouse, can enable the use of the Self-Localization System on the existing warehouses.

A demanda por mais flexibilidade nas fábricas e serviços originou um aumento no volume de operações internas de carga e descarga, devido à maior diversidade dos elementos transportados. Logo, na busca por um fluxo de materiais mais eficiente, as empresas passaram a investir em soluções tecnológicas, entre elas, o uso de Automated Guided Vehicles (AGVs), por conta do custo mais atrativo e do avanço em relação aos primeiros AGVs, que até então dependiam de uma infraestrutura adicional para suportar a navegação. Muitos AGVs modernos possuem movimentação livre e são orientados por sistemas que utilizam sensores para interpretar o ambiente, sendo assim, tornar os AGVs autônomos despertou o interesse de pesquisadores na área de robótica móvel para o desenvolvimento de sistemas capazes de auxiliar e coordenar a navegação. Novas técnicas de localização, tal como a localização baseada em marcadores reflexivos, e a construção de armazéns com layouts estruturados para a navegação viabilizaram o uso de AGVs autônomos, entretanto sua utilização em armazéns existentes ainda é um desafio. Neste contexto, o Laboratório de Robótica Móvel (LabRom) do Grupo de Mecatrônica da EESC/USP, através do projeto do Armazém Inteligente, tem pesquisado os problemas de: roteamento, gerenciamento das baterias, navegação e auto-localização. Robôs autônomos precisam de um sistema de auto-localização eficiente e preciso para navegar com segurança, o qual depende de um mapa e da interpretação do ambiente utilizando sensores embarcados. Para alcançar esse objetivo este trabalho propõe um Sistema de Auto-localização baseado no Extended Kalman Filter (EKF) como solução. O sistema, desenvolvido em linguagem C, interage com outros dois sistemas: roteamento e navegação e foi implementado em um armazém simulado utilizando o software Player/Stage, mostrando ser confiável no fornecimento de uma estimativa de localização baseada em odometria e landmarks com localização conhecida. O sistema foi novamente testado utilizando a odometria de um robô móvel Pioneer P3-AT e os valores de um sensor de medição laser 2D SICK LMS200 extraídos de um ambiente indoor real. Para este teste foi construído um feature-based map a partir de um desenho de planta baixa no formato CAD e utilizou-se o algoritmo de segmentação Iterative End-Point Fit (IEPF) para interpretar o ambiente. Os resultados mostraram que as vantagens oferecidas pelas características padronizadas de um ambiente indoor, semelhante a um armazém, podem viabilizar o uso do Sistema de Auto-localização em armazéns existentes. / The demand for more flexibility in factories and services led to an increase in the volume of internal operations of loading and unloading, due to the greater diversity of elements transported. Hence, in the search for a more efficient materials flow, companies went to invest in technology solutions, among them, the use of Automated Guided Vehicles (AGVs), on account of the more attractive cost and improvement over the first AGVs, which hitherto depended of an additional infrastructure to support navigation. Many modern AGVs have free movement and are guided by systems that use sensors to interpret the environment, thus make AGVs autonomous aroused the interest of researchers in the mobile robotics field to development of systems able to assist and coordinate the navigation. New localization techniques, such as localization based on reflective markers, and the construction of warehouses with structured layouts for navigation did feasible the use of autonomous AGVs, however its use in existing warehouses is still a challenge. In this context, the Mobile Robotics Lab (LabRom) of the Mechatronics Group of EESC/USP, through the Intelligent Warehouse Project, has researched the problems: routing, battery management, navigation and self-localization. Autonomous robots need an efficient and accurate self-localization system to safely navigate, which depends on one map and of the interpretation of the environment using embedded sensors. To achieve this goal, this work proposes a Self-Localization System based on the Extended Kalman Filter (EKF) as a solution. The system, developed in C language, interacts with two other systems: routing and navigation and was implemented in a simulated warehouse using the Player/Stage software, showing to be reliable in providing an estimative of localization based on odometry and landmarks with known localization. The system was again tested using the odometry of mobile robot Pioneer P3-AT and the values of a 2D Laser Rangefinder SICK LMS200 extracted from a real indoor environment. For this test was built a feature-based map from a floor plan design in CAD format and was used the segmentation algorithm Iterative End-Point Fit (IEPF) to interpret the environment. The results showed that the advantages offered by the standard features of indoor environment, like a warehouse, can enable the use of the Self-Localization System on the existing warehouses.

This dissertation aims to advance the field of structural optimization by creating and demonstrating new methodologies for the explicit inclusion of manufacturing issues. The case of composite aerospace structures was a main focus of this work as that field provides some of the greatest complexities in manufacturing yet also provides the greatest incentives to optimize structural performance. Firstly, the possibilities for modifying existing FEA based structural optimization methods to better capture manufacturing constraints are investigated. Examples of brick-based topology optimization, shell-based topology optimization, parametric sizing optimization and manufacturing process optimization are given. From these examples, a number of fundamental limitations to these methods were observed and are discussed. The key limitation that was uncovered related to a dichotomy between analytical methods (such as FEA) and CAD-type methods. Based on these observations, a new Knowledge-Based framework for structural optimization was suggested whereby manufacturing issues are integrally linked to the more conventional structural issues. A prototype system to implement this new framework was developed and is discussed. Finally, the validity of the framework was demonstrated by application to a generic composite rib design problem.

This dissertation studies the feature-based image comparison method and its application in Wireless Visual Sensor Networks. Wireless Visual Sensor Networks (WVSNs), formed by a large number of low-cost, small-size visual sensor nodes, represent a new trend in surveillance and monitoring practices. Although each single sensor has very limited capability in sensing, processing and transmission, by working together they can achieve various high level tasks. Sensor collaboration is essential to WVSNs and normally performed among sensors having similar measurements, which are called neighbor sensors. The directional sensing characteristics of imagers and the presence of visual occlusion present unique challenges to neighborhood formation, as geographically-close neighbors might not monitor similar scenes. Besides, the energy resource on the WVSNs is also very tight, with wireless communication and complicated computation consuming most energy in WVSNs. Therefore the feature-based image comparison method has been proposed, which directly compares the captured image from each visual sensor in an economical way in terms of both the computational cost and the transmission overhead. The feature-based image comparison method compares different images and aims to find similar image pairs using a set of local features from each image. The image feature is a numerical representation of the raw image and can be more compact in terms of the data volume than the raw image. The feature-based image comparison contains three steps: feature detection, descriptor calculation and feature comparison. For the step of feature detection, the dissertation proposes two computationally efficient corner detectors. The first detector is based on the Discrete Wavelet Transform that provides multi-scale corner point detection and the scale selection is achieved efficiently through a Gaussian convolution approach. The second detector is based on a linear unmixing model, which treats a corner point as the intersection of two or three “line” bases in a 3 by 3 region. The line bases are extracted through a constrained Nonnegative Matrix Factorization (NMF) approach and the corner detection is accomplished through counting the number of contributing bases in the linear mixture. For the step of descriptor calculation, the dissertation proposes an effective dimensionality reduction algorithm for the high dimensional Scale Invariant Feature Transform (SIFT) descriptors. A set of 40 SIFT descriptor bases are extracted through constrained NMF from a large training set and all SIFT descriptors are then projected onto the space spanned by these bases, achieving dimensionality reduction. The efficiency of the proposed corner detectors have been proven through theoretical analysis. In addition, the effectiveness of the proposed corner detectors and the dimensionality reduction approach has been validated through extensive comparison with several state-of-the-art feature detector/descriptor combinations.

The rapid advancement of information technology and its integration with the manufacturing technology increased the necessity of consistent and coherent data flow in the chain of Computer Aided Design (CAD)-Computer aided Manufacturing (CAM)-Computer Numerical Control (CNC). To achieve this, ISO 10303 standard (STEP), developed by ISO, is seen as a solution since STEP is independent of the environment on which design data, manufacturing data or machining data produced. In this thesis, efficiency of NC-code generation, with the inclusion of process planning data, from a STEP based CAD data is investigated. For the investigation purposes, software responsible for both building the STEP based CAD data and generating related NC-code automatically is developed. Using this software, several parts are designed / generated NC-codes are verified via CNC simulators and some test parts are produced. STEP AP224 based feature modeler, developed specifically for 2- axis rotational part design, includes / feature library, feature modeler employing SW2007 via API for visualization and preprocessor responsible for generation of STEP file in neutral format, called STEP Part 21. The NC-code generator includes / postprocessor responsible for STEP Part 21 interpretation, CNC machine tool and cutting tool database and preprocessor responsible for NC-code generation.

This thesis compares the state of the art methods and proposes solutions for human activity classification from video data. Human activity classification is finding the meaning of human activities, which are captured by the video. Classification of human activity is needed in order to improve surveillance video analysis and summarization, video data mining and robot intelligence. This thesis focuses on the classification of low level human activities which are used as an important information source to determine high level activities. In this study, the feature relation histogram based activity description proposed by Ryoo et al. (2009) is implemented and extended. The feature histogram is widely used in feature based approaches / however, the feature relation histogram has the ability to represent the locational information of the features. Our extension defines a new set of relations between the features, which makes the method more effective for action description. Classifications are performed and results are compared using feature histogram, Ryoo&rsquo / s feature relation histogram and our feature relation histogram using the same datasets and the feature type. Our experiments show that feature relation histogram performs slightly better than the feature histogram, our feature relation histogram is even better than both of the two. Although the difference is not clearly observable in the datasets containing periodic actions, a 12% improvement is observed for the non-periodic action datasets. Our work shows that the spatio-temporal relation represented by our new set of relations is a better way to represent the activity for classification.

Due to the ever-increasing volume of textual documents, text mining is a rapidly growing application of knowledge discovery in databases. Past text mining techniques predominately concentrated on discovering intra-document patterns from textual documents, such as text categorization, document clustering, query expansion, and event tracking. Mining inter-document patterns from textual documents has been largely ignored in the literature. This research focuses on discovering inter-document patterns, called evolution patterns, from document-sequences and proposed the evolution pattern discovery (EPD) technique for mining evolution patterns from a set of ordered sequences of documents. The discovery of evolution patterns can be applied in such domains as environmental scanning and knowledge management, and can be used to facilitate existing document management and retrieval techniques (e.g., event tracking).

Spectral imaging, a combination of spectroscopy and imaging, is a powerful tool for providing in situ material classification across a spatial scene. Typically spectral imaging analyses are interested in classification, though conventionally the classification is performed only after reconstruction of the spectral datacube, which can have upwards of 10⁹ signal elements. In this dissertation, I present a computational spectral imaging system, the Adaptive Feature-Specific Spectral Imaging Classifier (AFSSI-C), which yields direct classification across the spatial scene without reconstruction of the source datacube. With a dual disperser architecture and a programmable spatial light modulator which induces spectral filtering, the AFSSI-C measures specific projections of the spectral datacube which in turn feed an adaptive Bayesian classification and feature design framework. I present my work related to the design, construction, and testing of this instrument, which ultimately demonstrated significantly improved classification accuracy compared to legacy spectral imaging systems by first showing agreement with simulation, and then comparing to expected performance of traditional systems. As a result of its open aperture and adaptive filters, the AFSSI-C achieves 250 X better accuracy than pushbroom, whiskbroom, and tunable filter systems for a four-class problem at 0 dB TSNR (task signal-to-noise ratio) - a point where measurement noise is equal to the minimum separation between the library spectra. The AFSSI-C also achieves 100 X better accuracy than random projections at 0 dB TSNR.