Online Alpha Wave detector: an Embedded hardware-software implementation

The recent trend on embedded system development opens a new prospect for applications that in the past were not possible. The eye tracking for sleep and fatigue detection has become an important and useful application in industrial and automotive scenarios since fatigue is one of the most prevalent causes of earth-moving equipment accidents. Typical applications such as cameras, accelerometers and dermal analyzers are present on the market but have some inconvenient. This thesis project has used EEG signal, particularly, alpha waves, to overcome them by using an embedded software-hardware implementation to detect these signals in real time

Breaking Waves on the Ocean Surface

Thesis (Ph.D.)--University of Washington, 2016-08

Quantifying and Modeling the Influence of Forest on the Magnitude and Duration of Mountain Snow Storage in the Pacific Northwest, USA

Thesis (Ph.D.)--University of Washington, 2016-08

The Undersea World of the Sound Department: the Construction of Sonic Conventions in Sub-aqua Screen Environments

The availability of new underwater cameras and sub-aqua diving gear in the immediate post-war era opened up exciting possibilities for both narrative and documentary filmmakers. While the visual elements of this new world could now be more easily captured on film, the sound elements of the sub-aqua environment remained more elusive. What did, or should, this undersea world sound like? This article examines the use of sound in the sub-aqua scenes of both fictional and documentary films in the 1950s and asks questions about the methods used in the sonification of these worlds. Comparing the operation of underwater sound and human hearing with the production and post-production strategies used by filmmakers, I seek to identify the emergence of a sound convention and its implications for issues of cinematic realism. Central to this convention is the manipulation of sonic frequencies. The sound strategies adopted also raise questions about the malleability of viewer perspective and sound-image relationship in terms of a realist mode of address. Linked to this is the use of sound to enhance audience experience on an affective level. As well as underpinning cinematic realism, these new sound environments offered fresh experiences to audiences seeking new reasons to visit the cinema in an era of widening forms of entertainment.

Improving the Accuracy of 3-D Reconstruction in Robotic Vision Applications

Thesis (Ph.D.)--University of Washington, 2016-08

Exploring Forensic Nursing Ethics and Practice: Roles, Loyalties and Photodocumentation Practices

Thesis (Ph.D.)--University of Washington, 2016-08

Feature-based camera pose estimation

The estimating of the relative orientation and position of a camera is one of the integral topics in the field of computer vision. The accuracy of a certain Finnish technology company’s traffic sign inventory and localization process can be improved by utilizing the aforementioned concept. The company’s localization process uses video data produced by a vehicle installed camera. The accuracy of estimated traffic sign locations depends on the relative orientation between the camera and the vehicle. This thesis proposes a computer vision based software solution which can estimate a camera’s orientation relative to the movement direction of the vehicle by utilizing video data. The task was solved by using feature-based methods and open source software. When using simulated data sets, the camera orientation estimates had an absolute error of 0.31 degrees on average. The software solution can be integrated to be a part of the traffic sign localization pipeline of the company in question.

Dialectique des regards - l’autre et la construction du récit de soi dans l’auto-filmage pathographique

Le présent mémoire entend analyser la pratique de l’auto-filmage dans deux films : La Pudeur ou l’impudeur (Hervé Guibert) et Tarnation (Jonathan Caouette). Nous regroupons ces long-métrages sous l’étiquette « auto-filmage pathographique ». Le malade, s’émancipant de l’imagerie médicale et des pratiques cinématographiques institutionnelles, reprend l’image filmique à son compte, aidé en cela par une technologie toujours plus ergonomique. Cette résurgence de l’image du corps malade dans le champ social ne se fait pas sans heurt ; l’exposition de corps décharnés et agoniques convoque un imaginaire catastrophiste et contredit les rituels d’effacement du corps auxquels procède la société occidentale. La forme que prend le récit de soi dans l’auto-filmage pathographique dépend de la maladie qui affecte chaque créateur. Nous observons une redéfinition de la sincérité, en lien avec l’exercice autobiographique. Il s’agit d’utiliser, dans l’auto-filmage pathographique, certains procédés fictionnels pour créer un discours sur soi-même dont la véracité repose sur d’autres critères que ceux communément admis. L’auto-filmage pathographique suppose en ce sens un véritable changement d’attitude et la mise en place de techniques de soi. Il induit une forme de réconciliation avec sa propre identité physique et psychique. En cela, l’écriture filmique de soi est un agent transformateur de la vie et un exercice spirituel. Les réalisateurs ne sont cependant pas uniquement tournés vers eux-mêmes. Chacun inclut quelques privilégiés au coeur de sa démarche. Le soin de soi, dans l’auto-filmage pathographique, ne se désolidarise pas du soin des autres. Auto-filmage et caméra subjective entretiennent un lien dialectique qui donne son sens à l’auto-filmage pathographique et voit leur antagonisme éclater. L’individu s’auto-filmant n’est pas seul ; sa démarche n’est pas qu’un solipsisme. Elle se voit dépassée par l’émergence de l’autre dans le champ ou parfois même, sa prise en main de la caméra.

The Design and Implementation of a SD Card File System

SD card (Secure Digital Memory Card) is widely used in portable storage medium. Currently, latest researches on SD card, are mainly SD card controller based on FPGA (Field Programmable Gate Array). Most of them are relying on API interface (Application Programming Interface), AHB bus (Advanced High performance Bus), etc. They are dedicated to the realization of ultra high speed communication between SD card and upper systems. Studies about SD card controller, really play a vital role in the field of high speed cameras and other sub-areas of expertise. This design of FPGA-based file systems and SD2.0 IP (Intellectual Property core) does not only exhibit a nice transmission rate, but also achieve the systematic management of files, while retaining a strong portability and practicality. The file system design and implementation on a SD card covers the main three IP innovation points. First, the combination and integration of file system and SD card controller, makes the overall system highly integrated and practical. The popular SD2.0 protocol is implemented for communication channels. Pure digital logic design based on VHDL (Very-High-Speed Integrated Circuit Hardware Description Language), integrates the SD card controller in hardware layer and the FAT32 file system for the entire system. Secondly, the document management system mechanism makes document processing more convenient and easy. Especially for small files in batch processing, it can ease the pressure of upper system to frequently access and process them, thereby enhancing the overall efficiency of systems. Finally, digital design ensures the superior performance. For transmission security, CRC (Cyclic Redundancy Check) algorithm is for data transmission protection. Design of each module is platform-independent of macro cells, and keeps a better portability. Custom integrated instructions and interfaces may facilitate easily to use. Finally, the actual test went through multi-platform method, Xilinx and Altera FPGA developing platforms. The timing simulation and debugging of each module was covered. Finally, Test results show that the designed FPGA-based file system IP on SD card can support SD card, TF card and Micro SD with 2.0 protocols, and the successful implementation of systematic management for stored files, and supports SD bus mode. Data read and write rates in Kingston class10 card is approximately 24.27MB/s and 16.94MB/s.

Anomaly Detection in Noisy Images

Finding rare events in multidimensional data is an important detection problem that has applications in many fields, such as risk estimation in insurance industry, finance, flood prediction, medical diagnosis, quality assurance, security, or safety in transportation. The occurrence of such anomalies is so infrequent that there is usually not enough training data to learn an accurate statistical model of the anomaly class. In some cases, such events may have never been observed, so the only information that is available is a set of normal samples and an assumed pairwise similarity function. Such metric may only be known up to a certain number of unspecified parameters, which would either need to be learned from training data, or fixed by a domain expert. Sometimes, the anomalous condition may be formulated algebraically, such as a measure exceeding a predefined threshold, but nuisance variables may complicate the estimation of such a measure. Change detection methods used in time series analysis are not easily extendable to the multidimensional case, where discontinuities are not localized to a single point. On the other hand, in higher dimensions, data exhibits more complex interdependencies, and there is redundancy that could be exploited to adaptively model the normal data. In the first part of this dissertation, we review the theoretical framework for anomaly detection in images and previous anomaly detection work done in the context of crack detection and detection of anomalous components in railway tracks. In the second part, we propose new anomaly detection algorithms. The fact that curvilinear discontinuities in images are sparse with respect to the frame of shearlets, allows us to pose this anomaly detection problem as basis pursuit optimization. Therefore, we pose the problem of detecting curvilinear anomalies in noisy textured images as a blind source separation problem under sparsity constraints, and propose an iterative shrinkage algorithm to solve it. Taking advantage of the parallel nature of this algorithm, we describe how this method can be accelerated using graphical processing units (GPU). Then, we propose a new method for finding defective components on railway tracks using cameras mounted on a train. We describe how to extract features and use a combination of classifiers to solve this problem. Then, we scale anomaly detection to bigger datasets with complex interdependencies. We show that the anomaly detection problem naturally fits in the multitask learning framework. The first task consists of learning a compact representation of the good samples, while the second task consists of learning the anomaly detector. Using deep convolutional neural networks, we show that it is possible to train a deep model with a limited number of anomalous examples. In sequential detection problems, the presence of time-variant nuisance parameters affect the detection performance. In the last part of this dissertation, we present a method for adaptively estimating the threshold of sequential detectors using Extreme Value Theory on a Bayesian framework. Finally, conclusions on the results obtained are provided, followed by a discussion of possible future work.

LARGE MOTION VISUALIZATION AND ESTIMATION FOR FLAPPING WING SYSTEMS

Studies of fluid-structure interactions associated with flexible structures such as flapping wings require the capture and quantification of large motions of bodies that may be opaque. Motion capture of a free flying insect is considered by using three synchronized high-speed cameras. A solid finite element representation is used as a reference body and successive snapshots in time of the displacement fields are reconstructed via an optimization procedure. An objective function is formulated, and various shape difference definitions are considered. The proposed methodology is first studied for a synthetic case of a flexible cantilever structure undergoing large deformations, and then applied to a Manduca Sexta (hawkmoth) in free flight. The three-dimensional motions of this flapping system are reconstructed from image date collected by using three cameras. The complete deformation geometry of this system is analyzed. Finally, a computational investigation is carried out to understand the flow physics and aerodynamic performance by prescribing the body and wing motions in a fluid-body code. This thesis work contains one of the first set of such motion visualization and deformation analyses carried out for a hawkmoth in free flight. The tools and procedures used in this work are widely applicable to the studies of other flying animals with flexible wings as well as synthetic systems with flexible body elements.

High-resolution image reconstruction from multiple low-resolution images.

In this paper, we demonstrate a digital signal processing (DSP) algorithm for improving spatial resolution of images captured by CMOS cameras. The basic approach is to reconstruct a high resolution (HR) image from a shift-related low resolution (LR) image sequence. The aliasing relationship of Fourier transforms between discrete and continuous images in the frequency domain is used for mapping LR images to a HR image. The method of projection onto convex sets (POCS) is applied to trace the best estimate of pixel matching from the LR images to the reconstructed HR image. Computer simulations and preliminary experimental results have shown that the algorithm works effectively on the application of post-image-captured processing for CMOS cameras. It can also be applied to HR digital image reconstruction, where shift information of the LR image sequence is known.

Conversor de energia das ondas em energia elétrica com dispositivo de coluna de água oscilante: simulação numérica e estudo geométrico

Os oceanos representam um dos maiores recursos naturais, possuindo expressivo potencial energético, podendo suprir parte da demanda energética mundial. Nas últimas décadas, alguns dispositivos destinados à conversão da energia das ondas dos oceanos em energia elétrica têm sido estudados. No presente trabalho, o princípio de funcionamento do conversor do tipo Coluna de Água Oscilante, do inglês Oscillating Water Colum, (OWC) foi analisado numericamente. As ondas incidentes na câmara hidro-pneumática da OWC, causam um movimento alternado da coluna de água no interior da câmara, o qual produz um fluxo alternado de ar que passa pela chaminé. O ar passa e aciona uma turbina a qual transmite energia para um gerador elétrico. O objetivo do presente estudo foi investigar a influência de diferentes formas geométricas da câmara sobre o fluxo resultante de ar que passa pela turbina, que influencia no desempenho do dispositivo. Para isso, geometrias diferentes para o conversor foram analisadas empregando modelos computacionais 2D e 3D. Um modelo computacional desenvolvido nos softwares GAMBIT e FLUENT foi utilizado, em que o conversor OWC foi acoplado a um tanque de ondas. O método Volume of Fluid (VOF) e a teoria de 2ª ordem Stokes foram utilizados para gerar ondas regulares, permitindo uma interação mais realista entre o conversor, água, ar e OWC. O Método dos Volumes Finitos (MVF) foi utilizado para a discretização das equações governantes. Neste trabalho o Contructal Design (baseado na Teoria Constructal) foi aplicado pela primeira vez em estudos numéricos tridimensionais de OWC para fim de encontrar uma geometria que mais favorece o desempenho do dispositivo. A função objetivo foi a maximização da vazão mássica de ar que passa através da chaminé do dispositivo OWC, analisado através do método mínimos quadrados, do inglês Root Mean Square (RMS). Os resultados indicaram que a forma geométrica da câmara influencia na transformação da energia das ondas em energia elétrica. As geometrias das câmaras analisadas que apresentaram maior área da face de incidência das ondas (sendo altura constante), apresentaram também maior desempenho do conversor OWC. A melhor geometria, entre os casos desse estudo, ofereceu um ganho no desempenho do dispositivo em torno de 30% maior.

Leveraging Multiple Features for Image Retrieval and Matching

The goal of image retrieval and matching is to find and locate object instances in images from a large-scale image database. While visual features are abundant, how to combine them to improve performance by individual features remains a challenging task. In this work, we focus on leveraging multiple features for accurate and efficient image retrieval and matching. We first propose two graph-based approaches to rerank initially retrieved images for generic image retrieval. In the graph, vertices are images while edges are similarities between image pairs. Our first approach employs a mixture Markov model based on a random walk model on multiple graphs to fuse graphs. We introduce a probabilistic model to compute the importance of each feature for graph fusion under a naive Bayesian formulation, which requires statistics of similarities from a manually labeled dataset containing irrelevant images. To reduce human labeling, we further propose a fully unsupervised reranking algorithm based on a submodular objective function that can be efficiently optimized by greedy algorithm. By maximizing an information gain term over the graph, our submodular function favors a subset of database images that are similar to query images and resemble each other. The function also exploits the rank relationships of images from multiple ranked lists obtained by different features. We then study a more well-defined application, person re-identification, where the database contains labeled images of human bodies captured by multiple cameras. Re-identifications from multiple cameras are regarded as related tasks to exploit shared information. We apply a novel multi-task learning algorithm using both low level features and attributes. A low rank attribute embedding is joint learned within the multi-task learning formulation to embed original binary attributes to a continuous attribute space, where incorrect and incomplete attributes are rectified and recovered. To locate objects in images, we design an object detector based on object proposals and deep convolutional neural networks (CNN) in view of the emergence of deep networks. We improve a Fast RCNN framework and investigate two new strategies to detect objects accurately and efficiently: scale-dependent pooling (SDP) and cascaded rejection classifiers (CRC). The SDP improves detection accuracy by exploiting appropriate convolutional features depending on the scale of input object proposals. The CRC effectively utilizes convolutional features and greatly eliminates negative proposals in a cascaded manner, while maintaining a high recall for true objects. The two strategies together improve the detection accuracy and reduce the computational cost.