Video analysis based vehicle detection and tracking using an MCMC sampling framework

This article presents a probabilistic method for vehicle detection and tracking through the analysis of monocular images obtained from a vehicle-mounted camera. The method is designed to address the main shortcomings of traditional particle filtering approaches, namely Bayesian methods based on importance sampling, for use in traffic environments. These methods do not scale well when the dimensionality of the feature space grows, which creates significant limitations when tracking multiple objects. Alternatively, the proposed method is based on a Markov chain Monte Carlo (MCMC) approach, which allows efficient sampling of the feature space. The method involves important contributions in both the motion and the observation models of the tracker. Indeed, as opposed to particle filter-based tracking methods in the literature, which typically resort to observation models based on appearance or template matching, in this study a likelihood model that combines appearance analysis with information from motion parallax is introduced. Regarding the motion model, a new interaction treatment is defined based on Markov random fields (MRF) that allows for the handling of possible inter-dependencies in vehicle trajectories. As for vehicle detection, the method relies on a supervised classification stage using support vector machines (SVM). The contribution in this field is twofold. First, a new descriptor based on the analysis of gradient orientations in concentric rectangles is dened. This descriptor involves a much smaller feature space compared to traditional descriptors, which are too costly for real-time applications. Second, a new vehicle image database is generated to train the SVM and made public. The proposed vehicle detection and tracking method is proven to outperform existing methods and to successfully handle challenging situations in the test sequences.

Analysing written production competence descriptors for academic and professional purposes and their calibration to the CEFR

The Common European Framework of Reference for Languages (CEFR) "describes in a comprehensive way what language learners have to learn to do in order to use a language for communication and what knowledge and skills they have to develop so as to be able to act effectively" (Council of Europe, 2001: 1). This paper reports on the findings of two studies whose purpose was to assess written production competence descriptors meant for their inclusion into the Academic and Professional English Language Portfolio KELP) for students of engineering and architecture. The main objective of these studies was to establish whether the language competence descriptors were a satisfactory valid tool in their language programmes from the point of view of clarity, relevance and reliability, as perceived by the students and fellow English for Academic Purposes (RAP) / English for Science and Technology (EST) instructors. The studies shed light on how to improve unsatisfactory descriptors. Results show that the final descriptor lists were on the whole well calibrated and fairly well written: the great majority was considered valid for both teachers and students involved.

Comparación de algoritmos detectores de puntos singulares para reconocimiento de objetos en vídeo quirúrgico.

El análisis de vídeo laparoscópico ofrece nuevas posibilidades a la navegación quirúrgica al garantizar una incorporación mínima de tecnología en quirófano, evitando así alterar la ergonomía y los flujos de trabajo de las intervenciones. Una de sus principales ventajas es que puede servir como fuente de datos para reconstruir tridimensionalmente la escena laparoscópica, lo que permite dotar al cirujano de la sensación de profundidad perdida en este tipo de cirugía. En el presente trabajo de investigación se comparan dos detectores de puntos singulares, SIFT y SURF, para estimar cuál de los dos podría integrarse en un algoritmo de cálculo de coordenadas 3D, MonoSLAM, basado en la detección y el seguimiento de estos puntos singulares en los fotogramas del vídeo. Los resultados obtenidos posicionan a SURF como la mejor opción gracias a su rapidez y a su mayor capacidad de discriminación entre estructuras anatómicas e instrumental quirúrgico.

Identification procedure of mealy apples by instrumental means. Extraction of storage features.

Mealiness is a negative attribute of sensory texture that combines the sensation of a disaggregated tissue with the sensation of lack of juiciness. Since January 1996, a wide EC Project entitled : "Mealiness in fruits. Consumers perception and means for detection'" is being carried out. Within it, three sensory panels have been trained at : the Institute of Food Research (IFR, United Kingdom), the Instituto de Agroquímica y Tecnología de los Alimentos (IATA, Spain) and the Institut voor Agrotechnologisch Onderzoek (ATO-DLO, Netherlands) to assess mealiness in apples. In all three cases, mealiness has been described as a multidimensional sensory descriptor capable of gathering the loss of consistency (of crispness and of hardness) and of juiciness. Also within the EC Project several instrumental procedures have been tested for mealiness assessment. In this sense the Physical Properties Laboratory (ETS1A-UPM) has focused its aims in a first stage on performing instrumental tests for assessing some textural descriptors as crispiness, hardness and juiciness. The results obtained within these tests have shown to correlate well with the sensory measurements (Barreiro et Ruiz-Altisent, 1997) in apples, but also have succeed when trying to generate several texture degradation levels on peaches from which mealiness appears to be the last stage (Ortiz et al. 1997).

Direct estimation wave setup as a medium level in swash

The extreme runup is a key parameter for a shore risk analysis in which the accurate and quantitative estimation of the upper limit reached by waves is essential. Runup can be better approximated by splitting the setup and swash semi-amplitude contributions. In an experimental study recording setup becomes difficult due to infragravity motions within the surf zone, hence, it would be desirable to measure the setup with available methodologies and devices. In this research, an analysis is made of evaluated the convenience of direct estimation setup as the medium level in the swash zone for experimental runup analysis through a physical model. A physical mobile bed model was setup in a wave flume at the Laboratory for Maritime Experimentation of CEDEX. The wave flume is 36 metres long, 6.5 metres wide and 1.3 metres high. The physical model was designed to cover a reasonable range of parameters, three different slopes (1/50, 1/30 and 1/20), two sand grain sizes (D50 = 0.12 mm and 0.70 mm) and a range for the Iribarren number in deep water (ξ0) from 0.1 to 0.6. Best formulations were chosen for estimating a theoretical setup in the physical model application. Once theoretical setup had been obtained, a comparison was made with an estimation of the setup directly as a medium level of the oscillation in swash usually considered in extreme runup analyses. A good correlation was noted between both theoretical and time-averaging setup and a relation is proposed. Extreme runup is analysed through the sum of setup and semi-amplitude of swash. An equation is proposed that could be applied in strong foreshore slope-dependent reflective beaches.

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.

Image analysis by moment invariants using a set of step-like basis functions

Moment invariants have been thoroughly studied and repeatedly proposed as one of the most powerful tools for 2D shape identification. In this paper a set of such descriptors is proposed, being the basis functions discontinuous in a finite number of points. The goal of using discontinuous functions is to avoid the Gibbs phenomenon, and therefore to yield a better approximation capability for discontinuous signals, as images. Moreover, the proposed set of moments allows the definition of rotation invariants, being this the other main design concern. Translation and scale invariance are achieved by means of standard image normalization. Tests are conducted to evaluate the behavior of these descriptors in noisy environments, where images are corrupted with Gaussian noise up to different SNR values. Results are compared to those obtained using Zernike moments, showing that the proposed descriptor has the same performance in image retrieval tasks in noisy environments, but demanding much less computational power for every stage in the query chain.

New generation of human machine interfaces for controlling UAV through depth based gesture recognition

New forms of natural interactions between human operators and UAVs (Unmanned Aerial Vehicle) are demanded by the military industry to achieve a better balance of the UAV control and the burden of the human operator. In this work, a human machine interface (HMI) based on a novel gesture recognition system using depth imagery is proposed for the control of UAVs. Hand gesture recognition based on depth imagery is a promising approach for HMIs because it is more intuitive, natural, and non-intrusive than other alternatives using complex controllers. The proposed system is based on a Support Vector Machine (SVM) classifier that uses spatio-temporal depth descriptors as input features. The designed descriptor is based on a variation of the Local Binary Pattern (LBP) technique to efficiently work with depth video sequences. Other major consideration is the especial hand sign language used for the UAV control. A tradeoff between the use of natural hand signs and the minimization of the inter-sign interference has been established. Promising results have been achieved in a depth based database of hand gestures especially developed for the validation of the proposed system.

Traffic sign detection and tracking using robust 3D analysis

In this paper we present an innovative technique to tackle the problem of automatic road sign detection and tracking using an on-board stereo camera. It involves a continuous 3D analysis of the road sign during the whole tracking process. Firstly, a color and appearance based model is applied to generate road sign candidates in both stereo images. A sparse disparity map between the left and right images is then created for each candidate by using contour-based and SURF-based matching in the far and short range, respectively. Once the map has been computed, the correspondences are back-projected to generate a cloud of 3D points, and the best-fit plane is computed through RANSAC, ensuring robustness to outliers. Temporal consistency is enforced by means of a Kalman filter, which exploits the intrinsic smoothness of the 3D camera motion in traffic environments. Additionally, the estimation of the plane allows to correct deformations due to perspective, thus easing further sign classification.

Fast human pose estimation using 3D Zernike descriptors

Markerless video-based human pose estimation algorithms face a high-dimensional problem that is frequently broken down into several lower-dimensional ones by estimating the pose of each limb separately. However, in order to do so they need to reliably locate the torso, for which they typically rely on time coherence and tracking algorithms. Their losing track usually results in catastrophic failure of the process, requiring human intervention and thus precluding their usage in real-time applications. We propose a very fast rough pose estimation scheme based on global shape descriptors built on 3D Zernike moments. Using an articulated model that we configure in many poses, a large database of descriptor/pose pairs can be computed off-line. Thus, the only steps that must be done on-line are the extraction of the descriptors for each input volume and a search against the database to get the most likely poses. While the result of such process is not a fine pose estimation, it can be useful to help more sophisticated algorithms to regain track or make more educated guesses when creating new particles in particle-filter-based tracking schemes. We have achieved a performance of about ten fps on a single computer using a database of about one million entries.

Efficient hybrid monocular-stereo approach to on-board, video-based traffic sign detection and tracking

In this paper we propose an innovative method for the automatic detection and tracking of road traffic signs using an onboard stereo camera. It involves a combination of monocular and stereo analysis strategies to increase the reliability of the detections such that it can boost the performance of any traffic sign recognition scheme. Firstly, an adaptive color and appearance based detection is applied at single camera level to generate a set of traffic sign hypotheses. In turn, stereo information allows for sparse 3D reconstruction of potential traffic signs through a SURF-based matching strategy. Namely, the plane that best fits the cloud of 3D points traced back from feature matches is estimated using a RANSAC based approach to improve robustness to outliers. Temporal consistency of the 3D information is ensured through a Kalman-based tracking stage. This also allows for the generation of a predicted 3D traffic sign model, which is in turn used to enhance the previously mentioned color-based detector through a feedback loop, thus improving detection accuracy. The proposed solution has been tested with real sequences under several illumination conditions and in both urban areas and highways, achieving very high detection rates in challenging environments, including rapid motion and significant perspective distortion

Relation of Lagragian structures and drifter dynamics in the Gulf of Mexico

We use a Lagrangian descriptor (the so called function M) which measures the length of particle trajectories on the ocean surface over a given interval of time. With this tool we identify the Lagrangian skeleton of the flow and compare it on three datasets over the Gulf of Mexico during the year 2010. The satellite altimetry data used come from AVISO and simulations from HYCOM GOMl0.04 experiments 30.1 and 31.0. We contrast the Lagrangian structure and transport using the evolution of several surface drifters. We show that the agreement in relevant cases between Lagrangian structures and dynamics of drifters depends on the quality of the data on the studied area.

Measurements of different acoustic conditions in small rectangular rooms

Impulse response measurements are carried out in laboratory facilities at Ecophon, Sweden, simulating a typical classroom with varying suspended ceilings and furniture arrangements. The aim of these measurements is to build a reliable database of acoustical parameters in order to have enough data to validate the new acoustical simulation tool which is under development at Danmarks Tekniske Universitet, Denmark. The different classroom configurations are also simulated using ODEON Room Acoustics software and are compared with the measurements. The resulting information is essential for the development of the acoustical simulation tool because it will enable the elimination of prediction errors, especially those below the Schroeder frequency. The surface impedance of the materials used during the experiments is measured in a Kundt’s tube at DTU, in order to characterize them as accurately as possible at the time of incorporation into the model. A brief study about porous materials frequently used in classrooms is presented. Wide diferences are found between methods of measuring absorption coefficients and local or extended assumptions. RESUMEN. Mediciones de Respuesta al Impulso son llevadas a cabo en las instalaciones con que cuenta la empresa Ecophon en su sede central de Hyllinge, Suecia. En una de sus salas, se recrean diferentes configuraciones típicas de aula, variando la altura y composición de los techos, colocando paneles absorbentes de pared e incluyendo diferentes elementos mobiliario como pupitres y sillas. Tres diferentes materiales absorbentes porosos de 15, 20 y 50 mm de espesor, son utilizados como techos suspendidos así como uno de 40 mm es utilizado en forma de paneles. Todas las medidas son realizadas de acuerdo al estándar ISO 3382, utilizando 12 combinaciones de fuente sonora y micrófono para cada configuración, así como respetando las distancias entre ellos establecidas en la norma. El objetivo de toda esta serie de medidas es crear una base de datos de parámetros acústicos tales como tiempo de reverberación, índice de claridad o índice de inteligibilidad medidos bajo diferentes configuraciones con el objeto de que éstos sirvan de referencia para la validación de una nueva herramienta de simulación acústica llamada PARISM que está siendo desarrollada en este momento en la Danmarks Tekniske Universitet de Copenhague. Esta herramienta tendrá en cuenta la fase, tanto en propagación como en reflexión, así como el comportamiento angulodependiente de los materiales y la difusión producida por las superficies. Las diferentes configuraciones de aula recreadas en Hyllinge, son simuladas también utilizando el software de simulación acústica ODEON con el fin de establecer comparaciones entre medidas y simulaciones para discutir la validez de estas ultimas. La información resultante es esencial para el desarrollo de la nueva herramienta de simulación, especialmente los resultados por debajo de la frecuencia de corte de Schroeder, donde ODEON no produce predicciones precisas debido a que no tiene en cuenta la fase ni en propagación ni en reflexión. La impedancia de superficie de los materiales utilizados en los experimentos, todos ellos fabricados por la propia empresa Ecophon, es medida utilizando un tubo de Kundt. De este modo, los coeficientes de absorción de incidencia aleatoria son calculados e incorporados a las simulaciones. Además, estos coeficientes también son estimados mediante el modelo empírico de Miki, con el fin de ser comparados con los obtenidos mediante otros métodos. Un breve estudio comparativo entre coeficientes de absorción obtenidos por diversos métodos y el efecto producido por los materiales absorbentes sobre los tiempos de reverberación es realizado. Grandes diferencias son encontradas, especialmente entre los métodos de tubo de impedancia y cámara reverberante. La elección de reacción local o extendida a la hora de estimar los coeficientes también produce grandes diferencias entre los resultados. Pese a que la opción de absorción angular es activada en todas las simulaciones realizadas con ODEON para todos los materiales, los resultados son mucho más imprecisos de lo esperado a la hora de compararlos con los valores extraidos de las medidas de Respuesta al Impulso. En salas como las recreadas, donde una superficie es mucho más absorbente que las demás, las ondas sonoras tienden a incidir en la superficie altamente absorbente desde ángulos de incidencia muy pequeños. En este rango de ángulos de incidencia, las absorciones que presentan los materiales absorbentes porosos estudiados son muy pequeñas, pese a que sus valores de coeficientes de absorción de incidencia aleatoria son altos. Dado que como descriptor de las superficies en ODEON se utiliza el coeficiente de absorción de incidencia aleatoria, los tiempos de reverberación son siempre subestimados en las simulaciones, incluso con la opción de absorción angular activada. Esto es debido a que el algoritmo que ejecuta esta opción, solo tiene en cuenta el tamaño y posición de las superficies, mientras que el comportamiento angulodependiente es diferente para cada material. Es importante destacar, que cuando la opción es activada, los tiempos simulados se asemejan más a los medidos, por lo tanto esta característica sí produce ciertas mejoras pese a no modelar la angulodependencia perfectamente. Por otra parte, ODEON tampoco tiene en cuenta el fenómeno de difracción, ni acepta longitudes de superficie menores de una longitud de onda a frecuencias medias (30 cm) por lo que en las configuraciones que incluyen absorbentes de pared, los cuales presentan un grosor de 4 cm que no puede ser modelado, los tiempos de reverberación son siempre sobreestimados. Para evitar esta sobreestimación, diferentes métodos de correción son analizados. Todas estas deficiencias encontradas en el software ODEON, resaltan la necesidad de desarrollar cuanto antes la herramienta de simulación acústica PARISM, la cual será capaz de predecir el comportamiento del campo sonoro de manera precisa en este tipo de salas, sin incrementar excesivamente el tiempo de cálculo. En cuanto a los parámetros extraidos de las mediciones de Respuesta al Impulso, bajo ninguna de las configuraciones recreadas los tiempos de reverberación cumplen con las condiciones establecidas por la regulación danesa en materia de edificación. Es importante destacar que los experimentos son llevados a cabo en un edificio construido para uso industrial, en el que, pese a contar con un buen aislamiento acústico, los niveles de ruido pueden ser superiores a los existentes dentro del edificio donde finalmente se ubique el aula. Además, aunque algunos elementos de mobiliario como pupitres y sillas son incluidos, en una configuración real de aula normalmente aparecerían algunos otros como taquillas, que no solo presentarían una mayor absorción, sino que también dispersarían las ondas incidentes produciendo un mejor funcionamiento del techo absorbente. Esto es debido a que las ondas incidirían en el techo desde una mayor variedad de ángulos, y no solo desde ángulos cercanos a la dirección paralela al techo, para los cuales los materiales presentan absorciones muy bajas o casi nulas. En relación a los otros parámetros como índice de claridad o índice de inteligibilidad extraidos de las medidas, no se han podido extraer conclusiones válidas dada la falta de regulación existente. Sin embargo, el efecto que produce sobre ellos la inclusión de techos, paneles de pared y mobiliario sí es analizada, concluyendo que, como era de esperar, los mejores resultados son obtenidos cuando todos los elementos están presentes en la sala en el mismo momento.

Fast 2D to 3D conversion using a clustering-based hierarchical search in a machine learning framework

Automatic 2D-to-3D conversion is an important application for filling the gap between the increasing number of 3D displays and the still scant 3D content. However, existing approaches have an excessive computational cost that complicates its practical application. In this paper, a fast automatic 2D-to-3D conversion technique is proposed, which uses a machine learning framework to infer the 3D structure of a query color image from a training database with color and depth images. Assuming that photometrically similar images have analogous 3D structures, a depth map is estimated by searching the most similar color images in the database, and fusing the corresponding depth maps. Large databases are desirable to achieve better results, but the computational cost also increases. A clustering-based hierarchical search using compact SURF descriptors to characterize images is proposed to drastically reduce search times. A significant computational time improvement has been obtained regarding other state-of-the-art approaches, maintaining the quality results.

Study and analysis of the Leap Motion sensor and the Software Development Kit (SDK) for the implementation of visual Human Computer Interfaces

En este proyecto, se presenta un informe técnico sobre la cámara Leap Motion y el Software Development Kit correspondiente, el cual es un dispositivo con una cámara de profundidad orientada a interfaces hombre-máquina. Esto es realizado con el propósito de desarrollar una interfaz hombre-máquina basada en un sistema de reconocimiento de gestos de manos. Después de un exhaustivo estudio de la cámara Leap Motion, se han realizado diversos programas de ejemplo con la intención de verificar las capacidades descritas en el informe técnico, poniendo a prueba la Application Programming Interface y evaluando la precisión de las diferentes medidas obtenidas sobre los datos de la cámara. Finalmente, se desarrolla un prototipo de un sistema de reconocimiento de gestos. Los datos sobre la posición y orientación de la punta de los dedos obtenidos de la Leap Motion son usados para describir un gesto mediante un vector descriptor, el cual es enviado a una Máquina Vectores Soporte, utilizada como clasificador multi-clase.