Seleção de features guiada por atenção visual em imagens com fóvea

Visual attention is a very important task in autonomous robotics, but, because of its complexity, the processing time required is significant. We propose an architecture for feature selection using foveated images that is guided by visual attention tasks and that reduces the processing time required to perform these tasks. Our system can be applied in bottom-up or top-down visual attention. The foveated model determines which scales are to be used on the feature extraction algorithm. The system is able to discard features that are not extremely necessary for the tasks, thus, reducing the processing time. If the fovea is correctly placed, then it is possible to reduce the processing time without compromising the quality of the tasks outputs. The distance of the fovea from the object is also analyzed. If the visual system loses the tracking in top-down attention, basic strategies of fovea placement can be applied. Experiments have shown that it is possible to reduce up to 60% the processing time with this approach. To validate the method, we tested it with the feature algorithm known as Speeded Up Robust Features (SURF), one of the most efficient approaches for feature extraction. With the proposed architecture, we can accomplish real time requirements of robotics vision, mainly to be applied in autonomous robotics

Advancing Bag-of-visual-words Representations For Lesion Classification In Retinal Images.

Diabetic Retinopathy (DR) is a complication of diabetes that can lead to blindness if not readily discovered. Automated screening algorithms have the potential to improve identification of patients who need further medical attention. However, the identification of lesions must be accurate to be useful for clinical application. The bag-of-visual-words (BoVW) algorithm employs a maximum-margin classifier in a flexible framework that is able to detect the most common DR-related lesions such as microaneurysms, cotton-wool spots and hard exudates. BoVW allows to bypass the need for pre- and post-processing of the retinographic images, as well as the need of specific ad hoc techniques for identification of each type of lesion. An extensive evaluation of the BoVW model, using three large retinograph datasets (DR1, DR2 and Messidor) with different resolution and collected by different healthcare personnel, was performed. The results demonstrate that the BoVW classification approach can identify different lesions within an image without having to utilize different algorithms for each lesion reducing processing time and providing a more flexible diagnostic system. Our BoVW scheme is based on sparse low-level feature detection with a Speeded-Up Robust Features (SURF) local descriptor, and mid-level features based on semi-soft coding with max pooling. The best BoVW representation for retinal image classification was an area under the receiver operating characteristic curve (AUC-ROC) of 97.8% (exudates) and 93.5% (red lesions), applying a cross-dataset validation protocol. To assess the accuracy for detecting cases that require referral within one year, the sparse extraction technique associated with semi-soft coding and max pooling obtained an AUC of 94.2 ± 2.0%, outperforming current methods. Those results indicate that, for retinal image classification tasks in clinical practice, BoVW is equal and, in some instances, surpasses results obtained using dense detection (widely believed to be the best choice in many vision problems) for the low-level descriptors.

Navegação de robô móvel em ambiente com humanos

Trabalho Final de Mestrado para obtenção do grau de Mestre em Engenharia Mecânica

Localização de robôs móveis autônomos utilizando fusão sensorial de odometria e visão monocular

The development and refinement of techniques that make simultaneous localization and mapping (SLAM) for an autonomous mobile robot and the building of local 3-D maps from a sequence of images, is widely studied in scientific circles. This work presents a monocular visual SLAM technique based on extended Kalman filter, which uses features found in a sequence of images using the SURF descriptor (Speeded Up Robust Features) and determines which features can be used as marks by a technique based on delayed initialization from 3-D straight lines. For this, only the coordinates of the features found in the image and the intrinsic and extrinsic camera parameters are avaliable. Its possible to determine the position of the marks only on the availability of information of depth. Tests have shown that during the route, the mobile robot detects the presence of characteristics in the images and through a proposed technique for delayed initialization of marks, adds new marks to the state vector of the extended Kalman filter (EKF), after estimating the depth of features. With the estimated position of the marks, it was possible to estimate the updated position of the robot at each step, obtaining good results that demonstrate the effectiveness of monocular visual SLAM system proposed in this paper

Modelo para reconstrução 3D de cenas baseado em imagens

3D Reconstruction is the process used to obtain a detailed graphical model in three dimensions that represents some real objectified scene. This process uses sequences of images taken from the scene, so it can automatically extract the information about the depth of feature points. These points are then highlighted using some computational technique on the images that compose the used dataset. Using SURF feature points this work propose a model for obtaining depth information of feature points detected by the system. At the ending, the proposed system extract three important information from the images dataset: the 3D position for feature points; relative rotation and translation matrices between images; the realtion between the baseline for adjacent images and the 3D point accuracy error found.

Retinal slit lamp video mosaicking

Purpose To this day, the slit lamp remains the first tool used by an ophthalmologist to examine patient eyes. Imaging of the retina poses, however, a variety of problems, namely a shallow depth of focus, reflections from the optical system, a small field of view and non-uniform illumination. For ophthalmologists, the use of slit lamp images for documentation and analysis purposes, however, remains extremely challenging due to large image artifacts. For this reason, we propose an automatic retinal slit lamp video mosaicking, which enlarges the field of view and reduces amount of noise and reflections, thus enhancing image quality. Methods Our method is composed of three parts: (i) viable content segmentation, (ii) global registration and (iii) image blending. Frame content is segmented using gradient boosting with custom pixel-wise features. Speeded-up robust features are used for finding pair-wise translations between frames with robust random sample consensus estimation and graph-based simultaneous localization and mapping for global bundle adjustment. Foreground-aware blending based on feathering merges video frames into comprehensive mosaics. Results Foreground is segmented successfully with an area under the curve of the receiver operating characteristic curve of 0.9557. Mosaicking results and state-of-the-art methods were compared and rated by ophthalmologists showing a strong preference for a large field of view provided by our method. Conclusions The proposed method for global registration of retinal slit lamp images of the retina into comprehensive mosaics improves over state-of-the-art methods and is preferred qualitatively.

Content Based Video Retrieval using SURF Descriptor

This paper presents a Robust Content Based Video Retrieval (CBVR) system. This system retrieves similar videos based on a local feature descriptor called SURF (Speeded Up Robust Feature). The higher dimensionality of SURF like feature descriptors causes huge storage consumption during indexing of video information. To achieve a dimensionality reduction on the SURF feature descriptor, this system employs a stochastic dimensionality reduction method and thus provides a model data for the videos. On retrieval, the model data of the test clip is classified to its similar videos using a minimum distance classifier. The performance of this system is evaluated using two different minimum distance classifiers during the retrieval stage. The experimental analyses performed on the system shows that the system has a retrieval performance of 78%. This system also analyses the performance efficiency of the low dimensional SURF descriptor.

Intestinal content detection in capsule endoscopy using robust features

This work covers two aspects. First, it generally compares and summarizes the similarities and differences of state of the art feature detector and descriptor and second it presents a novel approach of detecting intestinal content (in particular bubbles) in capsule endoscopy images. Feature detectors and descriptors providing invariance to change of perspective, scale, signal-noise-ratio and lighting conditions are important and interesting topics in current research and the number of possible applications seems to be numberless. After analysing a selection of in the literature presented approaches, this work investigates in their suitability for applications information extraction in capsule endoscopy images. Eventually, a very good performing detector of intestinal content in capsule endoscopy images is presented. A accurate detection of intestinal content is crucial for all kinds of machine learning approaches and other analysis on capsule endoscopy studies because they occlude the field of view of the capsule camera and therefore those frames need to be excluded from analysis. As a so called “byproduct” of this investigation a graphical user interface supported Feature Analysis Tool is presented to execute and compare the discussed feature detectors and descriptor on arbitrary images, with configurable parameters and visualized their output. As well the presented bubble classifier is part of this tool and if a ground truth is available (or can also be generated using this tool) a detailed visualization of the validation result will be performed.

Neuroanatomic-based detection algorithm for automatic labeling of brain structures in brain injury

The number and grade of injured neuroanatomic structures and the type of injury determine the degree of impairment after a brain injury event and the recovery options of the patient. However, the body of knowledge and clinical intervention guides are basically focused on functional disorder and they still do not take into account the location of injuries. The prognostic value of location information is not known in detail either. This paper proposes a feature-based detection algorithm, named Neuroanatomic-Based Detection Algorithm (NBDA), based on SURF (Speeded Up Robust Feature) to label anatomical brain structures on cortical and sub-cortical areas. Themain goal is to register injured neuroanatomic structures to generate a database containing patient?s structural impairment profile. This kind of information permits to establish a relation with functional disorders and the prognostic evolution during neurorehabilitation procedures.

Benchmark Evaluation of HOG Descriptors as Features for Classification of Traffic Signs

The purpose of this paper is to analyze the performance of the Histograms of Oriented Gradients (HOG) as descriptors for traffic signs recognition. The test dataset consists of speed limit traffic signs because of their high inter-class similarities.   HOG features of speed limit signs, which were extracted from different traffic scenes, were computed and a Gentle AdaBoost classifier was invoked to evaluate the different features. The performance of HOG was tested with a dataset consisting of 1727 Swedish speed signs images. Different numbers of HOG features per descriptor, ranging from 36 features up 396 features, were computed for each traffic sign in the benchmark testing. The results show that HOG features perform high classification rate as the Gentle AdaBoost classification rate was 99.42%, and they are suitable to real time traffic sign recognition. However, it is found that changing the number of orientation bins has insignificant effect on the classification rate. In addition to this, HOG descriptors are not robust with respect to sign orientation.

Rate and processes of river network rearrangement during incipient faulting: The case of the Cahabon river, Guatemala

Deeply incised river networks are generally regarded as robust features that are not easily modified by erosion or tectonics. Although the reorganization of deeply incised drainage systems has been documented, the corresponding importance with regard to the overall landscape evolution of mountain ranges and the factors that permit such reorganizations are poorly understood. To address this problem, we have explored the rapid drainage reorganization that affected the Cahabon River in Guatemala during the Quaternary. Sediment-provenance analysis, field mapping, and electrical resistivity tomography (ERT) imaging are used to reconstruct the geometry of the valley before the river was captured. Dating of the abandoned valley sediments by the Be-10-Al-26 burial method and geomagnetic polarity analysis allow us to determine the age of the capture events and then to quantify several processes, such as the rate of tectonic deformation of the paleovalley, the rate of propagation of post-capture drainage reversal, and the rate at which canyons that formed at the capture sites have propagated along the paleovalley. Transtensional faulting started 1 to 3 million years ago, produced ground tilting and ground faulting along the Cahabon River, and thus generated differential uplift rate of 0.3 +/- 0.1 up to 0.7 +/- 0.4 mm . y(-1) along the river's course. The river responded to faulting by incising the areas of relative uplift and depositing a few tens of meters of sediment above the areas of relative subsidence. Then, the river experienced two captures and one avulsion between 700 ky and 100 ky. The captures breached high-standing ridges that separate the Cahabon River from its captors. Captures occurred at specific points where ridges are made permeable by fault damage zones and/or soluble rocks. Groundwater flow from the Cahabon River down to its captors likely increased the erosive power of the captors thus promoting focused erosion of the ridges. Valley-fill formation and capture occurred in close temporal succession, suggesting a genetic link between the two. We suggest that the aquifers accumulated within the valley-fills, increased the head along the subterraneous system connecting the Cahabon River to its captors, and promoted their development. Upon capture, the breached valley experienced widespread drainage reversal toward the capture sites. We attribute the generalized reversal to combined effects of groundwater sapping in the valley-fill, axial drainage obstruction by lateral fans, and tectonic tilting. Drainage reversal increased the size of the captured areas by a factor of 4 to 6. At the capture sites, 500 m deep canyons have been incised into the bedrock and are propagating upstream at a rate of 3 to 11 mm . y(-1) deepening at a rate of 0.7 to 1 5 mm . y(-1). At this rate, 1 to 2 million years will be necessary for headward erosion to completely erase the topographic expression of the paleovalley. It is concluded that the rapid reorganization of this drainage system was made possible by the way the river adjusted to the new tectonic strain field, which involved transient sedimentation along the river's course. If the river had escaped its early reorganization and had been given the time necessary to reach a new dynamic equilibrium, then the transient conditions that promoted capture would have vanished and its vulnerability to capture would have been strongly reduced.

Further analysis of open-respirometry systems: an a-compartmental mechanistic approach

A system is said to be "instantaneous" when for a given constant input an equilibrium output is obtained after a while. In the meantime, the output is changing from its initial value towards the equilibrium one. This is the transient period of the system and transients are important features of open-respirometry systems. During transients, one cannot compute the input amplitude directly from the output. The existing models (e.g., first or second order dynamics) cannot account for many of the features observed in real open-respirometry systems, such as time lag. Also, these models do not explain what should be expected when a system is speeded up or slowed down. The purpose of the present study was to develop a mechanistic approach to the dynamics of open-respirometry systems, employing basic thermodynamic concepts. It is demonstrated that all the main relevant features of the output dynamics are due to and can be adequately explained by a distribution of apparent velocities within the set of molecules travelling along the system. The importance of the rate at which the molecules leave the sensor is explored for the first time. The study approaches the difference in calibrating a system with a continuous input and with a "unit impulse": the former truly reveals the dynamics of the system while the latter represents the first derivative (in time) of the former and, thus, cannot adequately be employed in the apparent time-constant determination. Also, we demonstrate why the apparent order of the output changes with volume or flow.

Latin America and the Caribbean and the European Union: striving for a renewed partnership

Includes bibliography

La Plata en las vísperas montoneras : Una reconstrucción de las condiciones sociales y políticas de la masificación y radicalización política de la JP platense y su articulación con Montoneros (1970-72)

El trabajo trata sobre el proceso de radicalización política y masificación de Juventud Peronista de La Plata y su articulación con Montoneros. Abarcando el intenso período que puede extenderse desde mediados de 1970 hasta la vuelta de Perón a fines de 1972. Durante el mismo, la agrupación platense creció desarrollando una importante actividad política, sobre todo estudiantil y barrial, para formar parte hacia fines del '72 de la estrategia de masas de la organización Montoneros. En base a material periodístico y testimonial se busca reconstruir el 'escenario platense de la radicalización'. Es decir, por un lado las características demográficas, sociales, políticas, ideológicas locales que coadyuvaron a las movilizaciones. Por otra parte, se aborda la dinámica política. En primer término, la partidaria impulsada por la apertura electoral. Y luego, la revolucionaria acelerada por las organizaciones armadas. En ambas la JP platense se consolidó y terminó el proceso articulada con Montoneros para desarrollar posteriormente un importante 'trabajo de masas'

La Plata en las vísperas montoneras : Una reconstrucción de las condiciones sociales y políticas de la masificación y radicalización política de la JP platense y su articulación con Montoneros (1970-72)

El trabajo trata sobre el proceso de radicalización política y masificación de Juventud Peronista de La Plata y su articulación con Montoneros. Abarcando el intenso período que puede extenderse desde mediados de 1970 hasta la vuelta de Perón a fines de 1972. Durante el mismo, la agrupación platense creció desarrollando una importante actividad política, sobre todo estudiantil y barrial, para formar parte hacia fines del '72 de la estrategia de masas de la organización Montoneros. En base a material periodístico y testimonial se busca reconstruir el 'escenario platense de la radicalización'. Es decir, por un lado las características demográficas, sociales, políticas, ideológicas locales que coadyuvaron a las movilizaciones. Por otra parte, se aborda la dinámica política. En primer término, la partidaria impulsada por la apertura electoral. Y luego, la revolucionaria acelerada por las organizaciones armadas. En ambas la JP platense se consolidó y terminó el proceso articulada con Montoneros para desarrollar posteriormente un importante 'trabajo de masas'