Visual Tracking, Pose Estimation, and Control for Aerial Vehicles

El principal objetivo de esta tesis es dotar a los vehículos aéreos no tripulados (UAVs, por sus siglas en inglés) de una fuente de información adicional basada en visión. Esta fuente de información proviene de cámaras ubicadas a bordo de los vehículos o en el suelo. Con ella se busca que los UAVs realicen tareas de aterrizaje o inspección guiados por visión, especialmente en aquellas situaciones en las que no haya disponibilidad de estimar la posición del vehículo con base en GPS, cuando las estimaciones de GPS no tengan la suficiente precisión requerida por las tareas a realizar, o cuando restricciones de carga de pago impidan añadir sensores a bordo de los vehículos. Esta tesis trata con tres de las principales áreas de la visión por computador: seguimiento visual y estimación visual de la pose (posición y orientación), que a su vez constituyen la base de la tercera, denominada control servo visual, que en nuestra aplicación se enfoca en el empleo de información visual para controlar los UAVs. Al respecto, esta tesis se ocupa de presentar propuestas novedosas que permitan solucionar problemas relativos al seguimiento de objetos mediante cámaras ubicadas a bordo de los UAVs, se ocupa de la estimación de la pose de los UAVs basada en información visual obtenida por cámaras ubicadas en el suelo o a bordo, y también se ocupa de la aplicación de las técnicas propuestas para solucionar diferentes problemas, como aquellos concernientes al seguimiento visual para tareas de reabastecimiento autónomo en vuelo o al aterrizaje basado en visión, entre otros. Las diversas técnicas de visión por computador presentadas en esta tesis se proponen con el fin de solucionar dificultades que suelen presentarse cuando se realizan tareas basadas en visión con UAVs, como las relativas a la obtención, en tiempo real, de estimaciones robustas, o como problemas generados por vibraciones. Los algoritmos propuestos en esta tesis han sido probados con información de imágenes reales obtenidas realizando pruebas on-line y off-line. Diversos mecanismos de evaluación han sido empleados con el propósito de analizar el desempeño de los algoritmos propuestos, entre los que se incluyen datos simulados, imágenes de vuelos reales, estimaciones precisas de posición empleando el sistema VICON y comparaciones con algoritmos del estado del arte. Los resultados obtenidos indican que los algoritmos de visión por computador propuestos tienen un desempeño que es comparable e incluso mejor al de algoritmos que se encuentran en el estado del arte. Los algoritmos propuestos permiten la obtención de estimaciones robustas en tiempo real, lo cual permite su uso en tareas de control visual. El desempeño de estos algoritmos es apropiado para las exigencias de las distintas aplicaciones examinadas: reabastecimiento autónomo en vuelo, aterrizaje y estimación del estado del UAV. Abstract The main objective of this thesis is to provide Unmanned Aerial Vehicles (UAVs) with an additional vision-based source of information extracted by cameras located either on-board or on the ground, in order to allow UAVs to develop visually guided tasks, such as landing or inspection, especially in situations where GPS information is not available, where GPS-based position estimation is not accurate enough for the task to develop, or where payload restrictions do not allow the incorporation of additional sensors on-board. This thesis covers three of the main computer vision areas: visual tracking and visual pose estimation, which are the bases the third one called visual servoing, which, in this work, focuses on using visual information to control UAVs. In this sense, the thesis focuses on presenting novel solutions for solving the tracking problem of objects when using cameras on-board UAVs, on estimating the pose of the UAVs based on the visual information collected by cameras located either on the ground or on-board, and also focuses on applying these proposed techniques for solving different problems, such as visual tracking for aerial refuelling or vision-based landing, among others. The different computer vision techniques presented in this thesis are proposed to solve some of the frequently problems found when addressing vision-based tasks in UAVs, such as obtaining robust vision-based estimations at real-time frame rates, and problems caused by vibrations, or 3D motion. All the proposed algorithms have been tested with real-image data in on-line and off-line tests. Different evaluation mechanisms have been used to analyze the performance of the proposed algorithms, such as simulated data, images from real-flight tests, publicly available datasets, manually generated ground truth data, accurate position estimations using a VICON system and a robotic cell, and comparison with state of the art algorithms. Results show that the proposed computer vision algorithms obtain performances that are comparable to, or even better than, state of the art algorithms, obtaining robust estimations at real-time frame rates. This proves that the proposed techniques are fast enough for vision-based control tasks. Therefore, the performance of the proposed vision algorithms has shown to be of a standard appropriate to the different explored applications: aerial refuelling and landing, and state estimation. It is noteworthy that they have low computational overheads for vision systems.

A multi-resolution image alignment technique based on direct methods for pose estimation of aerial vehicles

In this paper, we seek to expand the use of direct methods in real-time applications by proposing a vision-based strategy for pose estimation of aerial vehicles. The vast majority of approaches make use of features to estimate motion. Conversely, the strategy we propose is based on a MR (Multi- Resolution) implementation of an image registration technique (Inverse Compositional Image Alignment ICIA) using direct methods. An on-board camera in a downwards-looking configuration, and the assumption of planar scenes, are the bases of the algorithm. The motion between frames (rotation and translation) is recovered by decomposing the frame-to-frame homography obtained by the ICIA algorithm applied to a patch that covers around the 80% of the image. When the visual estimation is required (e.g. GPS drop-out), this motion is integrated with the previous known estimation of the vehicles’ state, obtained from the on-board sensors (GPS/IMU), and the subsequent estimations are based only on the vision-based motion estimations. The proposed strategy is tested with real flight data in representative stages of a flight: cruise, landing, and take-off, being two of those stages considered critical: take-off and landing. The performance of the pose estimation strategy is analyzed by comparing it with the GPS/IMU estimations. Results show correlation between the visual estimation obtained with the MR-ICIA and the GPS/IMU data, that demonstrate that the visual estimation can be used to provide a good approximation of the vehicle’s state when it is required (e.g. GPS drop-outs). In terms of performance, the proposed strategy is able to maintain an estimation of the vehicle’s state for more than one minute, at real-time frame rates based, only on visual information.

Discrete loading of the foot during hurdle drills: A pilot investigation

Introduction: Lower limb function in hurdling is patently asymmetrical. The lead limb undertakes the preparatory and landing steps while the trail limb contends with the hurdle and recovery steps. Discrete loading profiles of these steps will reflect the asymmetrical function and may provide useful insight into injury mechanisms. A pilot study was undertaken to determine the loading profiles of the hurdle, landing and recovery steps of elite male hurdlers. Equivalent data for steps between hurdles, where the running action is more symmetrical, were used for the purpose of comparison, simultaneously minimising the confounding effect of speed. Methodology: In-shoe pressures were recorded (FScan, 200 Hz) for four elite male hurdlers while they completed a routine hurdle drill at a self-selected fast but sub-race speed. The drill comprised of three consecutive hurdles. Data for the hurdle, landing and recovery steps of the first and second hurdles, along with data for the running steps between hurdles 1 and 2, and 2 and 3, were used for the purpose of analysis. Peak pressures within 1cm2 masks were determined for the hallux, first, central and fifth metatarsals (T1, M1, M2–4 and M5 respectively). Peak pressure (kPa) and loading duration (ms) for the hurdle, landing and recovery steps are reported as a percentage of the respective limb-matched values for between-hurdle steps. Results/discussion: For between-hurdle steps, T1, M1 and M2–4 peak pressures were 312/357, 356/306 and 362/368 kPa, lead/trail limbs respectively. For the hurdle, landing and recovery steps, pressures at T1 and M1 increased. For T1 the increases were in the order of 17%, 36% and 8% (hurdle, landing and recovery steps, respectively) while the corresponding increases at M1 were 7%, 54% and 20%. Pressures at M2–4 were similar for all steps, while M5 loaded erratically. For the between-hurdle steps, the loading durations at T1, M1 and M2–4, were 160/162, 170/142 and 190/191 ms, respectively. For the landing step, loading duration decreased for T1, M1and M2–4 (−8%, −19% and −18%, respectively). In the hurdle step, loading duration decreased for the metatarsals but not for T1. Conclusions: The hurdling action leads to regional pressure increases that act for shorter durations in comparison to the between-hurdle running steps. These changes are most notable at the first metatarsal, a common site of foot injury.

The effect of recording situation on the echolocation calls of the New Zealand Lesser short-tailed bat (Mystacina tuberculata Gray).

Using a broad‐band recording system (150 Hz‐100 kHz) the echolocation calls of the lesser short‐tailed bat (Mystacina tuberculata) were recorded under three very different situations: free‐flying, flying within a flight cage, and on release from the hand. Calls of bats landing and feeding on a platform in Wellington Zoo were also recorded. Both the lowest frequency and frequency of peak amplitude of calls were significantly affected by the situation under which calls were recorded. Although the calls of free‐flying bats are different from those produced by bats foraging on the ground, it is unlikely that M. tuberculata uses echolocation to locate prey on the ground. No significant differences could be found between the calls emitted by male and female bats, and no consistent relationships were obvious between temporal and spectral call characteristics. There was some evidence to suggest that individual bats could be identified by their echolocation calls.

Impact of the California sea lion (Zalophus californianus) on salmon fisheries in Monterey Bay, California

To assess the impact of California sea lions (Zalophus californianus) on salmon fisheries in the Monterey Bay region of California, the percentages of hooked fish taken by sea lions in commercial and recreational salmon fisheries were estimated from 1997 to 1999. Onboard surveys of sea lion interactions with the commercial and recreational f isheries and dockside interviews with fishermen after their return to port were conducted in the ports of Santa Cruz, Moss Landing, and Monterey. Approximately 1745 hours of onboard and dockside surveys were conducted—924 hours in the commercial fishery and 821 hours in the recreational fishery (commercial passenger fishing vessels [CPFVs] and personal skiffs combined). Adult male California sea lions were responsible for 98.4% of the observed depredations of hooked salmon in the commercial and recreational fisheries in Monterey Bay. Mean annual percentages of hooked salmon taken by sea lions ranged from 8.5% to 28.6% in the commercial fishery, 2.2% to 18.36% in the CPFVs, and 4.0% to 17.5% in the personal skiff fishery. Depredation levels in the commercial and recreational salmon fisheries were greatest in 1998—likely a result of the large El Niño Southern Oscillation (ENSO) event that occurred from 1997 to 1998 that reduced natural prey resources. Commercial fishermen lost an estimated $18,031−$60,570 of gear and $225,833−$498,076 worth of salmon as a result of interactions with sea lions. Approximately 1.4−6.2% of the available salmon population was removed from the system as a result of sea lion interactions with the fishery. Assessing the impact of a growing sea lion population on fisheries stocks is difficult, but may be necessary for effective fisheries management.

Contribution of zooplankton to the fishery potential of some environments

Difference in zooplankton production in estuarine, coastal and oceanic realms of the Indian Ocean has been correlated to the fishery potential of the concerned area. In the estuarine habitat the estimated rate of secondary production can not sustain a fishery of the present magnitude. Direct correlation between fish landing and zooplankton is observed in the coastal waters. Compared to coastal waters zooplankton standing stock is low in the open ocean and this low production is compensated by a wider area. The estimated fishery potential of the ocean is much more than what is being exploited.

História da pescaria de camarāo de águas pouco produndas no Banco de Sofala

This work compiled data and information previously reported on the shrimp fishery and stock, since the beginning of commercial fishery in 1965. It includes a fishery description, the evolution of the fishing effort, landing and catch rates. The main historical management measures implemented in the fishery are also described, which includes the licensing regime, minimum mesh size requirements, closed seasons and total allowable catch. The methodologies used for stock assessment are also presented.

Vortex filament simulation of trailing vortex merger

Aircraft in high-lift configuration shed multiple vortices. These generally merge to form a downstream wake consisting of two counter-rotating vortices of equal strength. The understanding of the merger of two co-rotating trailing vortices is important in evaluating the separation criteria for different aircraft to prevent wake vortex hazards during landing and take-off. There is no existing theoretical method on the basis of which such norms can be set. The present study is aimed at gaining a better understanding of the behaviour of wake vortices behind the aircraft. Two dimensional studies are carried out using the vortex blob method and compared with Bertenyi's experiment. It is shown that inviscid two dimensional effects are insufficient to explain the observations. Three dimensional studies, using the vortex filament method, are applied to the same test case. Two Lamb-Oseen profile vortices of the same dimensions and initial separation as the experiment are allowed to evolve from a straight starting condition until a converged steady flow is achieved. The results obtained show good agreement with the experimental distance to merger. Core radius and separation behaviour is qualitatively similar to experiment, with the exception of rapid increases at first. This may be partially attributable to the choice of filament-element length, and recommended further work includes a convergence study for this parameter. Copyright © 2005 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

The low Reynolds number aerodynamics of leading edge flaps

Within the low Reynolds number regime at which birds and small air vehicles operate (Re=15,000-500,000), flow is beset with laminar separation bubbles and bubble burst which can lead to loss of lift and early onset of stall. Recent video footage of an eagle's wings in flight reveals an inconspicuous wing feature: the sudden deployment of a row of feathers from the lower surface of the wing to create a leading edge flap. An understanding of the aerodynamic function of this flap has been developed through a series of low speed wind tunnel tests performed on an Eppler E423 aerofoil. Experiments took place at Reynolds numbers ranging from 40000 to 140000 and angles of attack up to 30°. In the lower range of tested Reynolds numbers, application of the flap was found to substantially enhance aerofoil performance by augmenting the lift and limiting the drag at certain incidences. The leading edge flap was determined to act as a transition device at low Reynolds numbers, preventing the formation of separation bubbles and consequently decreasing the speed at which stall occurs during landing and manoeuvring.

地质导向钻井关键技术及方法

On the basis of analyzing the principle and realization of geo-steering drilling system, the key technologies and methods in it are systematically studied in this paper. In order to recognize lithology, distinguish stratum and track reservoirs, the techniques of MWD and data process about natural gamma, resistivity, inductive density and porosity are researched. The methods for pre-processing and standardizing MWD data and for converting geological data in directional and horizontal drilling are discussed, consequently the methods of data conversion between MD and TVD and those of formation description and adjacent well contrast are proposed. Researching the method of identifying sub-layer yields the techniques of single well explanation, multi-well evaluation and oil reservoir description. Using the extremum and variance clustering analysis realizes logging phase analysis and stratum subdivision and explanation, which provides a theoretical method and lays a technical basis for tracing oil reservoirs and achieving geo-steering drilling. Researching the technique for exploring the reservoir top with a holdup section provides a planning method of wellpath control scheme to trace oil and gas reservoir dynamically, which solves the problem of how to control well trajectory on condition that the layer’s TVD is uncertain. The control scheme and planning method of well path for meeting the demands of target hitting, soft landing and continuous steering respectively provide the technological guarantee to land safely and drill successfully for horizontal, extended-reach and multi-target wells. The integrative design and control technologies are researched based on geology, reservoir and drilling considering reservoir disclosing ratio as a primary index, and the methods for planning and control optimum wellpath under multi-target restriction, thus which lets the target wellpath lie the favorite position in oil reservoir during the process of geo-steering drilling. The BHA (bottomhole assembly) mechanical model is discussed using the finite element method, and the BHA design methods are given on the basis of mechanical analyses according to the shape of well trajectory and the characteristics of BHA’s structure and deformation. The methods for predicting the deflection rate of bent housing motors and designing their assemblies are proposed based on the principle of minimum potential energy, which can clearly show the relation between the BHA’s structure parameters and deflection rate, especially the key factors’ effect to the deflection rate. Moreover, the interaction model between bit and formation is discussed through the process of equivalent formation and equivalent bit considering the formation anisotropy and bit anisotropy on the basis of analyzing the influence factors of well trajectory. Accordingly, the inherence relationship among well trajectory, formation, bit and drilling direction is revealed, which lays the theory basis and technique for predicting and controlling well trajectory.

高斯束偏移的方法及应用

Gaussian beam is the asymptotic solution of wave equation concentred at the central ray. The Gaussian beam ray tracing method has many advantages over ray tracing method. Because of the prevalence of multipath and caustics in complex media, Kirchhoff migration usually can not get satisfactory images, but Gaussian beam migration can get better results.The Runge-Kutta method is used to carry out the raytracing, and the wavefront construction method is used to calculate the multipath wavefield. In this thesis, a new method to determine the starting point and initial direction of a new ray is proposed take advantage of the radius of curvature calculated by dynamic ray tracing method.The propagation characters of Gaussian beam in complex media are investigated. When Gaussian beam is used to calculate the Green function, the wave field near the source was decomposed in Gaussian beam in different direction, then the wave field at a point is the superposition of individual Gaussian beams.Migration aperture is the key factor for Kirchhoff migration. In this thesis, the criterion for the choice of optimum aperture is discussed taking advantage of stationary phase analysis. Two equivalent methods are proposed, but the second is more preferable.Gaussian beam migration based on dip scanning and its procedure are developed. Take advantage of the travel time, amplitude, and takeoff angle calculated by Gaussian beam method, the migration is accomplished.Using the proposed migration method, I carry out the numerical calculation of simple theoretical model, Marmousi model and field data, and compare the results with that of Kirchhoff migration. The comparison shows that the new Gaussian beam migration method can get a better result over Kirchhoff migration, with fewer migration noise and clearer image at complex structures.

Evaluación del stock norte – centro de la anchoveta peruana (Engraulis ringens Jenyns) por un modelo estadístico estructurado por edades

Se empleó un modelo poblacional estructurado por edades para estimar la abundancia, biomasa, biomasa desovante y el reclutamiento medio del stock norte – centro de la anchoveta peruana entre los años biológicos (octubre a setiembre) 1962-63 y 2007-08. El modelo, basado en un enfoque hacia adelante, fue optimizado minimizando las diferencias de los estimados del modelo y observaciones independientes de biomasa, desembarque y estructuras por edades de los desembarques. Los resultados muestran que han existido tres regímenes de productividad de dicho stock: el primero, entre 1962-63 y 1970-71, con la abundancia, biomasa, biomasa desovante y reclutamiento medio más altos; el segundo, entre 1971- 72 y 1990-91 con los niveles poblacionales más bajos; y el tercero, entre 1991-92 y 2007-08, con niveles intermedios. Parece claro que luego del colapso de las décadas de 1970 y 1980 el stock se ha recuperado de manera significativa aunque sin alcanzar los niveles de la década de 1960. Desde el año 2001-02 la biomasa desovante se ha mantenido por encima de cinco millones de toneladas, y la mortalidad por pesca ha mostrado una tendencia decreciente. Se demostró que el presente modelo estuvo en capacidad de captar la dinámica poblacional del stock norte – centro de la anchoveta validando su utilidad en las evaluaciones y monitoreo de la población de anchoveta.

Thermal Processing of Smoked Yellowfin Tuna (Thunnus albacares) in Flexible Pouches

This study is the first of its kind in India, where in smoked and thermal processed products have been developed using locally available wood as the source of wood smoke and flavoring and a shelf life of one year has been achieved. Retortable pouches of three layers, both imported and indigenous were found suitable to store thermal processed products. Heat penetration rate is quicker in retort pouches due to their thin profile in comparison to cans and hence the total process time is lesser. The nutritional and sensory attributes of the pouch products are better retained during processing. Hence these products are more acceptable than canned products. lndian vegetarian food products and fish curry products are available in the ready to eat form in the markets. Smoked and thermal processed products have not gained an entry to the market and hence this study will pave an opening for such products. Currently trade in tuna products from India is meager compared to the global trade. ln India proper utilization of tuna resources is yet to be achieved due to the lack of infrastructure for handling and knowledge of value addition. The raw material cost is also less due to the poor quality of the fish when landed. Hence, the availability of such products will help in the trade of tuna products, improving the quality of raw material landing and ultimately realizing a better value to the fishermen and processors.

Estudo da sinterabilidade de ligas de níquel obtidas por meio dos portadores de liga sic, si3n4 ou si metálico com grafita

Nickel alloys are frequently used in applications that require resistance at high temperatures associated with resistance to corrosion. Alloys of Ni-Si-C can be obtained by means of powder metallurgy in which powder mixtures are made of metallic nickel powders with additions of various alloying carriers for such were used in this study SiC, Si3N4 or Si metal with graphite. Carbonyl Ni powder with mean particle size of 11 mM were mixed with 3 wt% of SiC powders with an average particle size of 15, 30 and 50 μm and further samples were obtained containing 4 to 5% by mass of SiC with average particle size of 15 μm. Samples were also obtained by varying the carrier alloy, these being Si3N4 powder with graphite, with average particle size of 1.5 and 5 μm, respectively. As a metallic Si graphite with average particle size of 12.5 and 5 μm, respectively. The reference material used was nickel carbonyl sintered without adding carriers. Microstructural characterization of the alloys was made by optical microscopy and scanning electron microscopy with semi-quantitative chemical analysis. We determined the densities of the samples and measurement of microhardness. We studied the dissociation of carriers alloy after sintering at 1200 ° C for 60 minutes. Was evaluated also in the same sintering conditions, the influence of the variation of average particle size of the SiC carrier to the proportion of 3% by mass. Finally, we studied the influence of variation of the temperatures of sintering at 950, 1080 and 1200 ° C without landing and also with heights of 30, 60, 120 and 240 minutes for sintering where the temperature was 950 °C. Dilatometry curves showed that the SiC sintered Ni favors more effectively than other carriers alloy analyzed. SiC with average particle size of 15 μm active sintering the alloy more effectively than other SiC used. However, with the chemical and morphological analyzes for all leagues, it was observed that there was dissociation of SiC and Si3N4, as well as diffusion of Si in Ni matrix and carbon cluster and dispersed in the matrix, which also occurred for the alloys with Si carriers and metallic graphite. So the league that was presented better results containing Si Ni with graphite metallic alloy as carriers, since this had dispersed graphite best in the league, reaching the microstructural model proposed, which is necessary for material characteristic of solid lubricant, so how we got the best results when the density and hardness of the alloy

Caracterização do regime alimentar de Oreochromis niloticus (Linnaeus, 1758) e Tilapia rendalli (Boulenger, 1897) na represa de Barra Bonita, Médio Rio Tietê, SP

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)