Human activity recognition with accelerometry: novel time and frequency features

Human Activity Recognition systems require objective and reliable methods that can be used in the daily routine and must offer consistent results according with the performed activities. These systems are under development and offer objective and personalized support for several applications such as the healthcare area. This thesis aims to create a framework for human activities recognition based on accelerometry signals. Some new features and techniques inspired in the audio recognition methodology are introduced in this work, namely Log Scale Power Bandwidth and the Markov Models application. The Forward Feature Selection was adopted as the feature selection algorithm in order to improve the clustering performances and limit the computational demands. This method selects the most suitable set of features for activities recognition in accelerometry from a 423th dimensional feature vector. Several Machine Learning algorithms were applied to the used accelerometry databases – FCHA and PAMAP databases - and these showed promising results in activities recognition. The developed algorithm set constitutes a mighty contribution for the development of reliable evaluation methods of movement disorders for diagnosis and treatment applications.

Hand detection on images based on deformable part models and additional features

Hand detection on images has important applications on person activities recognition. This thesis focuses on PASCAL Visual Object Classes (VOC) system for hand detection. VOC has become a popular system for object detection, based on twenty common objects, and has been released with a successful deformable parts model in VOC2007. A hand detection on an image is made when the system gets a bounding box which overlaps with at least 50% of any ground truth bounding box for a hand on the image. The initial average precision of this detector is around 0.215 compared with a state-of-art of 0.104; however, color and frequency features for detected bounding boxes contain important information for re-scoring, and the average precision can be improved to 0.218 with these features. Results show that these features help on getting higher precision for low recall, even though the average precision is similar.

Alteración del centro de gravedad y riesgo de caída en pacientes con trastornos neurológicos. Bogotá. 2014

Introducción: el objetivo de la rehabilitación en los pacientes neurológicos es reducir su discapacidad, potencializar su funcionalidad y promover su independencia para así permitirle desempeñar un rol activo en la comunidad. Por esta razón, es de gran importancia prevenir la aparición de eventos adversos en esta población mediante la detección temprana de factores de riesgo que conllevan a pacientes con enfermedad cerebro vascular a presentar caídas y a su vez le permita a los profesionales de la salud generar estrategias para minimizar su incidencia, complicaciones y/o secuelas. Objetivo: determinar los factores relacionados con la alteración del centro de gravedad y riesgos de caída en paciente con trastornos neurológicos Metodología: se realizó un estudio cuantitativo descriptivo exploratorio retrospectivo donde se tuvieron en cuenta pacientes diagnóstico de enfermedades neurológicas que presentaron o no antecedentes de caídas junto con otros criterios de inclusión y exclusión, tomados de las bases de datos de la institución Mobility Group. Resultados: Se evaluaron en total 19 sujetos (52,6 % hombres) con edad media de 48,37 años. Se clasificaron en dos grupos (alto y bajo riesgo de caída ) de pacientes según la aplicación de la escala de riesgo de caída de J.H. Downton a los cuales se aplicó estadística descriptiva para describir su comportamiento en referencia a factores como diagnóstico clínico, presencia de dolor, fuerza y tono de miembros inferiores en los principales grupos musculares , clasificación de espasticidad y finalmente con la alteración de la trayectoria del centro de gravedad según la valoración realizada con el dispositivo THERA-TRAINER BALANCE.

Automatic segmentation of latent fingerprints

Latent fingerprints are routinely found at crime scenes due to the inadvertent contact of the criminals' finger tips with various objects. As such, they have been used as crucial evidence for identifying and convicting criminals by law enforcement agencies. However, compared to plain and rolled prints, latent fingerprints usually have poor quality of ridge impressions with small fingerprint area, and contain large overlap between the foreground area (friction ridge pattern) and structured or random noise in the background. Accordingly, latent fingerprint segmentation is a difficult problem. In this paper, we propose a latent fingerprint segmentation algorithm whose goal is to separate the fingerprint region (region of interest) from background. Our algorithm utilizes both ridge orientation and frequency features. The orientation tensor is used to obtain the symmetric patterns of fingerprint ridge orientation, and local Fourier analysis method is used to estimate the local ridge frequency of the latent fingerprint. Candidate fingerprint (foreground) regions are obtained for each feature type; an intersection of regions from orientation and frequency features localizes the true latent fingerprint regions. To verify the viability of the proposed segmentation algorithm, we evaluated the segmentation results in two aspects: a comparison with the ground truth foreground and matching performance based on segmented region. © 2012 IEEE.

Um método para melhoria de qualidade de imagens médicas utilizando a transformada wavelet

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

A Quantitative analysis of pulsed signals emitted by wild bottlenose dolphins

Common bottlenose dolphins (Tursiops truncatus), produce a wide variety of vocal emissions for communication and echolocation, of which the pulsed repertoire has been the most difficult to categorize. Packets of high repetition, broadband pulses are still largely reported under a general designation of burst-pulses, and traditional attempts to classify these emissions rely mainly in their aural characteristics and in graphical aspects of spectrograms. Here, we present a quantitative analysis of pulsed signals emitted by wild bottlenose dolphins, in the Sado estuary, Portugal (2011-2014), and test the reliability of a traditional classification approach. Acoustic parameters (minimum frequency, maximum frequency, peak frequency, duration, repetition rate and inter-click-interval) were extracted from 930 pulsed signals, previously categorized using a traditional approach. Discriminant function analysis revealed a high reliability of the traditional classification approach (93.5% of pulsed signals were consistently assigned to their aurally based categories). According to the discriminant function analysis (Wilk's Λ = 0.11, F3, 2.41 = 282.75, P < 0.001), repetition rate is the feature that best enables the discrimination of different pulsed signals (structure coefficient = 0.98). Classification using hierarchical cluster analysis led to a similar categorization pattern: two main signal types with distinct magnitudes of repetition rate were clustered into five groups. The pulsed signals, here described, present significant differences in their time-frequency features, especially repetition rate (P < 0.001), inter-click-interval (P < 0.001) and duration (P < 0.001). We document the occurrence of a distinct signal type-short burst-pulses, and highlight the existence of a diverse repertoire of pulsed vocalizations emitted in graded sequences. The use of quantitative analysis of pulsed signals is essential to improve classifications and to better assess the contexts of emission, geographic variation and the functional significance of pulsed signals.

Relative frequency, clinical, neuroimaging, and postsurgical features of pediatric temporal lobe epilepsy

We describe the relative frequency, clinical features, neuroimaging and pathological results, and outcome after pharmacological or surgical intervention for a series of pediatric patients with temporal lobe epilepsy (TLE) from an epilepsy center in Brazil. The medical records of children younger than 12 years with features strongly suggestive of TLE were reviewed from January 1999 to June 1999. Selected children were evaluated regarding clinical, EEG, and magnetic resonance imaging (MRI) investigation and divided into three groups according to MRI: group 1 (G1, N = 9), patients with hippocampal atrophy; group 2 (G2, N = 10), patients with normal MRI, and group 3 (G3, N = 12), patients with other specific temporal lesions. A review of 1732 records of children with epilepsy revealed 31 cases with TLE (relative frequency of 1.79%). However, when the investigation was narrowed to cases with intractable seizures that needed video-EEG monitoring (N = 68) or epilepsy surgery (N = 32), the relative frequency of TLE increased to 19.11 (13/68) and 31.25% (10/32), respectively. At the beginning of the study, 25 of 31 patients had a high seizure frequency (80.6%), which declined to 11 of 31 (35.5%) at the conclusion of the study, as a consequence of pharmacological and/or surgical therapy. This improvement in seizure control was significant in G1 (P < 0.05) and G3 (P < 0.01) mainly due to good postsurgical outcome, and was not significant in G2 (P > 0.1, McNemar's test). These results indicate that the relative frequency of TLE in children was low, but increased considerably among cases with pharmacoresistant seizures. Patients with specific lesions were likely to undergo surgery, with good postoperative outcomes.

A virtual experiment for the detection of specific features in the frequency response of a coupled nonlinear and linear oscillator

This paper presents an investigation into some practical issues that may be present in a real experiment, when trying to validate the theoretical frequency response curve of a two degree-of-freedom nonlinear system consisting of coupled linear and nonlinear oscillators. Some specific features, such as detached resonance curves, have been theoretically predicted in multi degree-of-freedom nonlinear oscillators, when subject to harmonic excitation, and the system parameters have been shown to be fundamental in achieving such features. When based on a simplified model, approximate analytical expression for the frequency response curves may be derived, which may be validated by the numerical solutions. In a real experiment, however, the practical achievability of such features was previously shown to be greatly affected by small disturbances induced by gravity and inertia, which led to some solutions becoming unstable which had been predicted to be stable. In this work a practical system configuration is proposed where such effects are reduced so that the previous limitations are overcome. A virtual experiment is carried out where a detailed multi-body model of the oscillator is assembled and the effects on the system response are investigated.

High frequency components of the AE signal as useful features for bond/fibre break discriminators

Abstract not available

Radial variation of anatomical features, wood density and decay resistance in teak (Tectona grandis) from two qualities of growing sites and two climatic regions of Costa Rica

The objective of this study was to show the radial variation of some anatomic characteristics, wood density and natural durability of teak (Tectona grandis L.F.) growing in Costa Rica. Samples of trees 13 years old were obtained from two growing sites (high and low growing) of plantations established in a humid tropical climate (CHT) and dry tropical climate (CST). The variables measured of the fibers as well as for the rays were not affected by the climate or the type of growing site, except for the length of the fibers. The fibers of teak wood from the best growing site were significantly larger. Vessels were found with a greater frequency for the CST but mostly solitary in comparison with the CBT. Average density, maximum density and the variation within the ring presented a light higher magnitude for the CST. The quality of the growing site did not affect these variables. The resistance of fungus attack was similar in the area of heartwood near the pith compared to the heartwood near the sapwood for all the conditions evaluated. Nevertheless, it was observed in some trees a similar resistance of fungus attack for areas of sapwood compared to similar areas of heartwood.

Continuous high-frequency turbulence and suspended sediment concentration measurements in an upper estuary

The present study details new turbulence field measurements conducted continuously at high frequency for 50 hours in the upper zone of a small subtropical estuary with semi-diurnal tides. Acoustic Doppler velocimetry was used, and the signal was post-processed thoroughly. The suspended sediment concentration wad further deduced from the acoustic backscatter intensity. The field data set demonstrated some unique flow features of the upstream estuarine zone, including some low-frequency longitudinal oscillations induced by internal and external resonance. A striking feature of the data set is the large fluctuations in all turbulence properties and suspended sediment concentration during the tidal cycle. This feature has been rarely documented.

Estimating population features of the anomuran crab Petrolisthes armatus (Porcellanidae) in a remaining and impacted mangrove area of the western Atlantic

This study evaluated biological aspects of Petrolisthes armatus inhabiting a remaining but disturbed mangrove area of Brazil. Samples were taken from March 2005 to July 2006, during low tide in the rocky-shore region. The size frequency distribution for all the individuals collected was bimodal. The sex ratio obtained was not different from 1: 1, and the population was classified as standard. A higher reproductive ratio and earlier ovigerous conditions were found in relation to another southern population; these can be hypothesized as adaptations to life in a stressed environment. Even though the population living in Ara has been subject to an environment frequently disturbed by human-produced pollutants, our results show no sign of negative effects on reproductive stages, recruits or members of the population in general. The population profiles of P. armatus show some peculiarities when compared to other populations inhabiting non disturbed environments.

Metallic Voice: Physiological and Acoustic Features

The metallic voice is usually confused with ring or nasality by singers and nontrained listeners. who are not used to perceptual vocal analysis. They believe a metallic voice results from a rise in fundamental frequency. A diagnostic error in this aspect may lead to lowering pitch, an incorrect procedure that Could Cause vocal overload and fatigue. The purpose of this article is to Study the quality of metallic voice considering the correlation between information of the physiological and acoustic plans, based on a perceptive consensual assumption. Fiberscopic video pharyngolaryngoscopy was performed on 21 professional singers while speaking vowel [e]-in normal and metallic modes to observe muscular movements and structural changes of the velopharynx, pharynx, and larynx. Vocal samples captured simultaneously to the fiberscopic examination were acoustically analyzed. Frequency and amplitude of the first four formants (F(1), F(2), F(3), and F(4)) were extracted by means of linear predictor coefficients (LPC) Spectrum and were statistically analyzed. Vocal tract adjustments such as velar lowering, pharyngeal wall narrowing, laryngeal rise, aryepiglottic, and lateral laryngeal constrictions were frequently found: there were no significant changes in frequency and amplitude of F(1) in the metallic voiced there were significant increases in amplitudes of F(2), F(3), and F(4) and in frequency for F, metallic Voice perceived as louder was correlated to an increase ill amplitude of F(3) and F(4). Physiological adjustments of velopharynx, pharynx, and larynx are combined in characterizing the metallic voice and can be acoustically related to changes in formant pattern.

High frequency of lipoprotein risk levels for cardiovascular disease in Takayasu arteritis

The objective of this study is to characterize the lipoprotein risk levels in Takayasu arteritis (TA) patients and its possible association with disease activity and glucocorticoid use. Twenty-five female TA patients were consecutively included and compared with 30 age-, gender-, and body mass index-matched healthy controls. Demographic features and the lipid profile were determined and cardiovascular risk levels were evaluated according to NCEP/ATPIII. Total cholesterol (TC), LDL-c, HDL-c, and triglycerides were determined after a 12-h overnight fast. Exclusion criteria were conditions that interfere in the lipid profile. The disease duration was 6.6 +/- 7.4 years; 30% had clinical activity and 80% had laboratory activity. Regarding NCEP/ATPIII risk levels, TA patients presented higher frequency of lipid risk compared to controls: high TC (48% vs. 20%, p = 0.04), low HDL-c (20% vs. 0%, p = 0.015), and high triglycerides (36% vs. 10%, p = 0.026). No difference was observed related to LDL-c risk levels between both groups (40% vs. 20%, p = 0.14). Remarkably, 60% of the patients had at least one lipid risk factor for cardiovascular disease. No difference in the lipids was observed between patients with and without clinical activity; however, those with laboratory activity showed lower levels of HDL-c (1.37 +/- 0.42 vs. 2.00 +/- 0.63 mmol/L, p = 0.012) than patients without this activity. A negative correlation was found between HDL-c and CRP levels (r = -0.42, p = 0.04). Lipids were similar in patients under glucocorticoid compared to those without this therapy. This is the first study to identify that TA, an inflammatory disease, has a proatherogenic lipid profile which is associated to laboratory disease activity.