Processing and rendering of Fourier domain optical coherence tomography images at a line rate over 524 kHz using a graphics processing unit

In Fourier domain optical coherence tomography (FD-OCT), a large amount of interference data needs to be resampled from the wavelength domain to the wavenumber domain prior to Fourier transformation. We present an approach to optimize this data processing, using a graphics processing unit (GPU) and parallel processing algorithms. We demonstrate an increased processing and rendering rate over that previously reported by using GPU paged memory to render data in the GPU rather than copying back to the CPU. This avoids unnecessary and slow data transfer, enabling a processing and display rate of well over 524,000 A-scan/s for a single frame. To the best of our knowledge this is the fastest processing demonstrated to date and the first time that FD-OCT processing and rendering has been demonstrated entirely on a GPU.

Area-delay Trade-offs of Texture Decompressors for a Graphics Processing Unit

Graphics Processing Units have become a booster for the microelectronics industry. However, due to intellectual property issues, there is a serious lack of information on implementation details of the hardware architecture that is behind GPUs. For instance, the way texture is handled and decompressed in a GPU to reduce bandwidth usage has never been dealt with in depth from a hardware point of view. This work addresses a comparative study on the hardware implementation of different texture decompression algorithms for both conventional (PCs and video game consoles) and mobile platforms. Circuit synthesis is performed targeting both a reconfigurable hardware platform and a 90nm standard cell library. Area-delay trade-offs have been extensively analyzed, which allows us to compare the complexity of decompressors and thus determine suitability of algorithms for systems with limited hardware resources.

Parallel hyperspectral unmixing method via split augmented lagrangian on GPU

One of the main problems of hyperspectral data analysis is the presence of mixed pixels due to the low spatial resolution of such images. Linear spectral unmixing aims at inferring pure spectral signatures and their fractions at each pixel of the scene. The huge data volumes acquired by hyperspectral sensors put stringent requirements on processing and unmixing methods. This letter proposes an efficient implementation of the method called simplex identification via split augmented Lagrangian (SISAL) which exploits the graphics processing unit (GPU) architecture at low level using Compute Unified Device Architecture. SISAL aims to identify the endmembers of a scene, i.e., is able to unmix hyperspectral data sets in which the pure pixel assumption is violated. The proposed implementation is performed in a pixel-by-pixel fashion using coalesced accesses to memory and exploiting shared memory to store temporary data. Furthermore, the kernels have been optimized to minimize the threads divergence, therefore achieving high GPU occupancy. The experimental results obtained for the simulated and real hyperspectral data sets reveal speedups up to 49 times, which demonstrates that the GPU implementation can significantly accelerate the method's execution over big data sets while maintaining the methods accuracy.

Reconstrução de imagem médica de mamografia por emissão de positrões (PEM) com GPU

Dissertação para obtenção do Grau de Mestre em Engenharia Biomédica

Algorithmic skeleton framework for the orchestration of GPU computations

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Graphics Processor Unit Based Parallelization of Optimized Baseline Wander Filtering Algorithms for Long-term Electrocardiography

Long-term electrocardiogram (ECG) often suffers from relevant noise. Baseline wander in particular is pronounced in ECG recordings using dry or esophageal electrodes, which are dedicated for prolonged registration. While analog high-pass filters introduce phase distortions, reliable offline filtering of the baseline wander implies a computational burden that has to be put in relation to the increase in signal-to-baseline ratio (SBR). Here we present a graphics processor unit (GPU) based parallelization method to speed up offline baseline wander filter algorithms, namely the wavelet, finite, and infinite impulse response, moving mean, and moving median filter. Individual filter parameters were optimized with respect to the SBR increase based on ECGs from the Physionet database superimposed to auto-regressive modeled, real baseline wander. A Monte-Carlo simulation showed that for low input SBR the moving median filter outperforms any other method but negatively affects ECG wave detection. In contrast, the infinite impulse response filter is preferred in case of high input SBR. However, the parallelized wavelet filter is processed 500 and 4 times faster than these two algorithms on the GPU, respectively, and offers superior baseline wander suppression in low SBR situations. Using a signal segment of 64 mega samples that is filtered as entire unit, wavelet filtering of a 7-day high-resolution ECG is computed within less than 3 seconds. Taking the high filtering speed into account, the GPU wavelet filter is the most efficient method to remove baseline wander present in long-term ECGs, with which computational burden can be strongly reduced.

3D simulation of complex shading affecting PV systems taking benefit from the power of graphics cards developed for the video game industry

Shading reduces the power output of a photovoltaic (PV) system. The design engineering of PV systems requires modeling and evaluating shading losses. Some PV systems are affected by complex shading scenes whose resulting PV energy losses are very difficult to evaluate with current modeling tools. Several specialized PV design and simulation software include the possibility to evaluate shading losses. They generally possess a Graphical User Interface (GUI) through which the user can draw a 3D shading scene, and then evaluate its corresponding PV energy losses. The complexity of the objects that these tools can handle is relatively limited. We have created a software solution, 3DPV, which allows evaluating the energy losses induced by complex 3D scenes on PV generators. The 3D objects can be imported from specialized 3D modeling software or from a 3D object library. The shadows cast by this 3D scene on the PV generator are then directly evaluated from the Graphics Processing Unit (GPU). Thanks to the recent development of GPUs for the video game industry, the shadows can be evaluated with a very high spatial resolution that reaches well beyond the PV cell level, in very short calculation times. A PV simulation model then translates the geometrical shading into PV energy output losses. 3DPV has been implemented using WebGL, which allows it to run directly from a Web browser, without requiring any local installation from the user. This also allows taken full benefits from the information already available from Internet, such as the 3D object libraries. This contribution describes, step by step, the method that allows 3DPV to evaluate the PV energy losses caused by complex shading. We then illustrate the results of this methodology to several application cases that are encountered in the world of PV systems design. Keywords: 3D, modeling, simulation, GPU, shading, losses, shadow mapping, solar, photovoltaic, PV, WebGL

Parallel Computer Vision Algorithms for Graphics Processing Units

La evolución de los teléfonos móviles inteligentes, dotados de cámaras digitales, está provocando una creciente demanda de aplicaciones cada vez más complejas que necesitan algoritmos de visión artificial en tiempo real; puesto que el tamaño de las señales de vídeo no hace sino aumentar y en cambio el rendimiento de los procesadores de un solo núcleo se ha estancado, los nuevos algoritmos que se diseñen para visión artificial han de ser paralelos para poder ejecutarse en múltiples procesadores y ser computacionalmente escalables. Una de las clases de procesadores más interesantes en la actualidad se encuentra en las tarjetas gráficas (GPU), que son dispositivos que ofrecen un alto grado de paralelismo, un excelente rendimiento numérico y una creciente versatilidad, lo que los hace interesantes para llevar a cabo computación científica. En esta tesis se exploran dos aplicaciones de visión artificial que revisten una gran complejidad computacional y no pueden ser ejecutadas en tiempo real empleando procesadores tradicionales. En cambio, como se demuestra en esta tesis, la paralelización de las distintas subtareas y su implementación sobre una GPU arrojan los resultados deseados de ejecución con tasas de refresco interactivas. Asimismo, se propone una técnica para la evaluación rápida de funciones de complejidad arbitraria especialmente indicada para su uso en una GPU. En primer lugar se estudia la aplicación de técnicas de síntesis de imágenes virtuales a partir de únicamente dos cámaras lejanas y no paralelas—en contraste con la configuración habitual en TV 3D de cámaras cercanas y paralelas—con información de color y profundidad. Empleando filtros de mediana modificados para la elaboración de un mapa de profundidad virtual y proyecciones inversas, se comprueba que estas técnicas son adecuadas para una libre elección del punto de vista. Además, se demuestra que la codificación de la información de profundidad con respecto a un sistema de referencia global es sumamente perjudicial y debería ser evitada. Por otro lado se propone un sistema de detección de objetos móviles basado en técnicas de estimación de densidad con funciones locales. Este tipo de técnicas es muy adecuada para el modelado de escenas complejas con fondos multimodales, pero ha recibido poco uso debido a su gran complejidad computacional. El sistema propuesto, implementado en tiempo real sobre una GPU, incluye propuestas para la estimación dinámica de los anchos de banda de las funciones locales, actualización selectiva del modelo de fondo, actualización de la posición de las muestras de referencia del modelo de primer plano empleando un filtro de partículas multirregión y selección automática de regiones de interés para reducir el coste computacional. Los resultados, evaluados sobre diversas bases de datos y comparados con otros algoritmos del estado del arte, demuestran la gran versatilidad y calidad de la propuesta. Finalmente se propone un método para la aproximación de funciones arbitrarias empleando funciones continuas lineales a tramos, especialmente indicada para su implementación en una GPU mediante el uso de las unidades de filtraje de texturas, normalmente no utilizadas para cómputo numérico. La propuesta incluye un riguroso análisis matemático del error cometido en la aproximación en función del número de muestras empleadas, así como un método para la obtención de una partición cuasióptima del dominio de la función para minimizar el error. ABSTRACT The evolution of smartphones, all equipped with digital cameras, is driving a growing demand for ever more complex applications that need to rely on real-time computer vision algorithms. However, video signals are only increasing in size, whereas the performance of single-core processors has somewhat stagnated in the past few years. Consequently, new computer vision algorithms will need to be parallel to run on multiple processors and be computationally scalable. One of the most promising classes of processors nowadays can be found in graphics processing units (GPU). These are devices offering a high parallelism degree, excellent numerical performance and increasing versatility, which makes them interesting to run scientific computations. In this thesis, we explore two computer vision applications with a high computational complexity that precludes them from running in real time on traditional uniprocessors. However, we show that by parallelizing subtasks and implementing them on a GPU, both applications attain their goals of running at interactive frame rates. In addition, we propose a technique for fast evaluation of arbitrarily complex functions, specially designed for GPU implementation. First, we explore the application of depth-image–based rendering techniques to the unusual configuration of two convergent, wide baseline cameras, in contrast to the usual configuration used in 3D TV, which are narrow baseline, parallel cameras. By using a backward mapping approach with a depth inpainting scheme based on median filters, we show that these techniques are adequate for free viewpoint video applications. In addition, we show that referring depth information to a global reference system is ill-advised and should be avoided. Then, we propose a background subtraction system based on kernel density estimation techniques. These techniques are very adequate for modelling complex scenes featuring multimodal backgrounds, but have not been so popular due to their huge computational and memory complexity. The proposed system, implemented in real time on a GPU, features novel proposals for dynamic kernel bandwidth estimation for the background model, selective update of the background model, update of the position of reference samples of the foreground model using a multi-region particle filter, and automatic selection of regions of interest to reduce computational cost. The results, evaluated on several databases and compared to other state-of-the-art algorithms, demonstrate the high quality and versatility of our proposal. Finally, we propose a general method for the approximation of arbitrarily complex functions using continuous piecewise linear functions, specially formulated for GPU implementation by leveraging their texture filtering units, normally unused for numerical computation. Our proposal features a rigorous mathematical analysis of the approximation error in function of the number of samples, as well as a method to obtain a suboptimal partition of the domain of the function to minimize approximation error.

Advances in characterisation, calibration and data processing speed of optical coherence tomography systems

This thesis describes advances in the characterisation, calibration and data processing of optical coherence tomography (OCT) systems. Femtosecond (fs) laser inscription was used for producing OCT-phantoms. Transparent materials are generally inert to infra-red radiations, but with fs lasers material modification occurs via non-linear processes when the highly focused light source interacts with the materials. This modification is confined to the focal volume and is highly reproducible. In order to select the best inscription parameters, combination of different inscription parameters were tested, using three fs laser systems, with different operating properties, on a variety of materials. This facilitated the understanding of the key characteristics of the produced structures with the aim of producing viable OCT-phantoms. Finally, OCT-phantoms were successfully designed and fabricated in fused silica. The use of these phantoms to characterise many properties (resolution, distortion, sensitivity decay, scan linearity) of an OCT system was demonstrated. Quantitative methods were developed to support the characterisation of an OCT system collecting images from phantoms and also to improve the quality of the OCT images. Characterisation methods include the measurement of the spatially variant resolution (point spread function (PSF) and modulation transfer function (MTF)), sensitivity and distortion. Processing of OCT data is a computer intensive process. Standard central processing unit (CPU) based processing might take several minutes to a few hours to process acquired data, thus data processing is a significant bottleneck. An alternative choice is to use expensive hardware-based processing such as field programmable gate arrays (FPGAs). However, recently graphics processing unit (GPU) based data processing methods have been developed to minimize this data processing and rendering time. These processing techniques include standard-processing methods which includes a set of algorithms to process the raw data (interference) obtained by the detector and generate A-scans. The work presented here describes accelerated data processing and post processing techniques for OCT systems. The GPU based processing developed, during the PhD, was later implemented into a custom built Fourier domain optical coherence tomography (FD-OCT) system. This system currently processes and renders data in real time. Processing throughput of this system is currently limited by the camera capture rate. OCTphantoms have been heavily used for the qualitative characterization and adjustment/ fine tuning of the operating conditions of OCT system. Currently, investigations are under way to characterize OCT systems using our phantoms. The work presented in this thesis demonstrate several novel techniques of fabricating OCT-phantoms and accelerating OCT data processing using GPUs. In the process of developing phantoms and quantitative methods, a thorough understanding and practical knowledge of OCT and fs laser processing systems was developed. This understanding leads to several novel pieces of research that are not only relevant to OCT but have broader importance. For example, extensive understanding of the properties of fs inscribed structures will be useful in other photonic application such as making of phase mask, wave guides and microfluidic channels. Acceleration of data processing with GPUs is also useful in other fields.

Parallel hyperspectral unmixing on GPUs

This letter presents a new parallel method for hyperspectral unmixing composed by the efficient combination of two popular methods: vertex component analysis (VCA) and sparse unmixing by variable splitting and augmented Lagrangian (SUNSAL). First, VCA extracts the endmember signatures, and then, SUNSAL is used to estimate the abundance fractions. Both techniques are highly parallelizable, which significantly reduces the computing time. A design for the commodity graphics processing units of the two methods is presented and evaluated. Experimental results obtained for simulated and real hyperspectral data sets reveal speedups up to 100 times, which grants real-time response required by many remotely sensed hyperspectral applications.

Estrutura a partir do Movimento em Sistemas Autónomos

Mestrado em Engenharia Electrotécnica e de Computadores - Ramo de Sistemas Autónomos

Parallel method for sparse semisupervised hyperspectral unmixing

Parallel hyperspectral unmixing problem is considered in this paper. A semisupervised approach is developed under the linear mixture model, where the abundance's physical constraints are taken into account. The proposed approach relies on the increasing availability of spectral libraries of materials measured on the ground instead of resorting to endmember extraction methods. Since Libraries are potentially very large and hyperspectral datasets are of high dimensionality a parallel implementation in a pixel-by-pixel fashion is derived to properly exploits the graphics processing units (GPU) architecture at low level, thus taking full advantage of the computational power of GPUs. Experimental results obtained for real hyperspectral datasets reveal significant speedup factors, up to 164 times, with regards to optimized serial implementation.

Acceleration of physics simulation engine through OpenCL

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Heterogeneous computing with an algorithmic skeleton framework

The Graphics Processing Unit (GPU) is present in almost every modern day personal computer. Despite its specific purpose design, they have been increasingly used for general computations with very good results. Hence, there is a growing effort from the community to seamlessly integrate this kind of devices in everyday computing. However, to fully exploit the potential of a system comprising GPUs and CPUs, these devices should be presented to the programmer as a single platform. The efficient combination of the power of CPU and GPU devices is highly dependent on each device’s characteristics, resulting in platform specific applications that cannot be ported to different systems. Also, the most efficient work balance among devices is highly dependable on the computations to be performed and respective data sizes. In this work, we propose a solution for heterogeneous environments based on the abstraction level provided by algorithmic skeletons. Our goal is to take full advantage of the power of all CPU and GPU devices present in a system, without the need for different kernel implementations nor explicit work-distribution.To that end, we extended Marrow, an algorithmic skeleton framework for multi-GPUs, to support CPU computations and efficiently balance the work-load between devices. Our approach is based on an offline training execution that identifies the ideal work balance and platform configurations for a given application and input data size. The evaluation of this work shows that the combination of CPU and GPU devices can significantly boost the performance of our benchmarks in the tested environments, when compared to GPU-only executions.