Dissertations / Theses on the topic 'Motion'

This thesis describes a number of studies conducted to examine three different facets of horizontal motion processing in the auditory system. Firstly, when a sound moved around a stationary listener (“source motion”); secondly, when subjects engaged in head rotations while sources remained stationary (“self motion”) and lastly when subjects engaged in self motion during simultaneous source motion. Previous studies in the field have explored these issues separately, and much remains unknown. For “source motion”, a localisation based “snapshot” psychophysical model remains the most commonly used narrative in describing this process, given the lack of clarity about the neural pathways underlying motion perception. However, it remains unclear whether (or how) such a framework can generalise to different stimulus conditions. For “self motion”, studies reported here have considered the sensory implications of head motion in the presence of a stationary sound, questioning how auditory spatial perception remains stable and exploring the perceptual benefits from dynamic localisation cues. Yet, the underlying interactions between audition and the head motor plant remain unclear, particularly at the faster head turn velocities. Lastly, there is a scarcity of studies probing the how listeners perceive a moving source during simultaneous self motion, even though it encapsulates concepts in both self and source motion, providing a unique opportunity to help frame our understanding of the sensorimotor mechanisms involved. We addressed these questions with three psychophysical experiments, and proposed a leaky integrative framework as an alternative to the “snapshot” model.

In this dissertation, we propose a universal motion-based control framework that supports general functionalities on 2D and 3D user interfaces with a single integrated design. We develop a hybrid framework of optical and inertial sensing technologies to track 6-DOF (degrees of freedom) motion of a handheld device, which includes the explicit 6-DOF (position and orientation in the global coordinates) and the implicit 6-DOF (acceleration and angular speed in the device-wise coordinates). Motion recognition is another key function of the universal motion-based control and contains two parts: motion gesture recognition and air-handwriting recognition. The interaction technique of each task is carefully designed to follow a consistent mental model and ensure the usability. The universal motion-based control achieves seamless integration of 2D and 3D interactions, motion gestures, and air-handwriting. Motion recognition by itself is a challenging problem. For motion gesture recognition, we propose a normalization procedure to effectively address the large in-class motion variations among users. The main contribution is the investigation of the relative effectiveness of various feature dimensions (of tracking signals) for motion gesture recognition in both user-dependent and user-independent cases. For air-handwriting recognition, we first develop a strategy to model air-handwriting with basic elements of characters and ligatures. Then, we build word-based and letter-based decoding word networks for air-handwriting recognition. Moreover, we investigate the detection and recognition of air-fingerwriting as an extension to air-handwriting. To complete the evaluation of air-handwriting, we conduct usability study to support that air-handwriting is suitable for text input on a motion-based user interface.

Motion capture technology is recognised as a standard tool in the computer animation pipeline. It provides detailed movement for animators; however, it also introduces problems and brings concerns for creating realistic and convincing motion for character animation. In this thesis, the post-processing techniques are investigated that result in realistic motion generation. Anumber of techniques are introduced that are able to improve the quality of generated motion from motion capture data, especially when integrating motion transitions from different motion clips. The presented motion data reconstruction technique is able to build convincing realistic transitions from existing motion database, and overcome the inconsistencies introduced by traditional motion blending techniques. It also provides a method for animators to re-use motion data more efficiently. Along with the development of motion data transition reconstruction, the motion capture data mapping technique was investigated for skeletal movement estimation. The per-frame based method provides animators with a real-time and accurate solution for a key post-processing technique. Although motion capture systems capture physically-based motion for character animation, no physical information is included in the motion capture data file. Using the knowledge of biomechanics and robotics, the relevant information for the captured performer are able to be abstracted and a mathematical-physical model are able to be constructed; such information is then applied for physics-based motion data correction whenever the motion data is edited.

Studiens syfte var att undersöka hur man inleder fysiska tester och träning efter hjärtinfarkt samt

hur det påverkar konditionen och vilka faktorer som förstärker effekten av träning. Resultatet av

den empiriska undersökning avsåg att jämföras med resultatet av den forskning som andra

forskare utfört. Studien beskriver hjärtinfarkten vad gäller anatomi och patologi samt ger en

överblick över olika forskningsresultat inom området. En pilotstudie med en testperson

innehållande ett förtest, en träningsperiod och ett eftertest genomfördes.

Det stora antalet tester som gjorts visar sitt tydliga språk. Fysisk träning efter hjärtinfarkt är

gynnsamt och förbättrar återhämtningen avsevärt. Träningen ger förbättrad livskvalite i form av

större ork och uthållighet i vardagslivet. Desto mångsidigare motionsform kombinerat med

avslappnings- och andningsteknikövningar desto bättre resultat. Även den utförda pilotstudien

pekar åt samma håll.

Complex realistic human motion sequences satisfying environmental constraints can be created by motion capture, which is a reliable way to reproduce human motions. However, motion capture data is difficult to modify in order to obtain variant motion sequences for multiple tasks. In this thesis, a system for synthesizing complex realistic human motion sequences based on a small group of sample motions to satisfy constraints is proposed. Methods are proposed for the system to preprocesses raw motion capture data to create sample motions that can be easily modified for the purpose of meeting specific requirements, while maintaining the subtleties of the origin motion capture data. Methods for the system to scan user-input constraints, to choose the best sample motion and synthesize the motion sequence based on route affected by the constraint are also proposed. Each generated motion piece is blended with the default motion, and thus a motion sequence composed of several pieces of motion based on constraints will be generated. Artifacts that arise during motion generation are identified and handled properly. Experimental results will show that the system can create cyclical sample motions from motion capture data, generate motion pieces based on environmental constraints, and synthesize complex realistic human motion sequences.

When the relative velocity between the different objects in a scene and the camera is relative large---compared with the camera's exposure time---in the resulting image we have a distortion called motion blur. In the past, a lot of algorithms have been proposed for estimating the relative velocity from one or, most of the time, more images. The motion blur is generally considered an extra source of noise and is eliminated, or is assumed nonexistent. Unlike most of these approaches, it is feasible to estimate the Optical Flow map using only the information encoded in the motion blur. This thesis presents an algorithm that estimates the velocity vector of an image patch using the motion blur only, in two steps. The information used for the estimation of the velocity vectors is extracted from the frequency domain, and the most computationally expensive operation is the Fast Fourier Transform that transforms the image from the spatial to the frequency domain. Consequently, the complexity of the algorithm is bound by this operation into O( nlog(n)). The first step consists of using the response of a family of steerable filters applied on the log of the Power Spectrum in order to calculate the orientation of the velocity vector. The second step uses a technique called Cepstral Analysis. More precisely, the log power spectrum is treated as another signal and we examine the Inverse Fourier Transform of it in order to estimate the magnitude of the velocity vector. Experiments have been conducted on artificially blurred images and with real world data, and an error analysis on these results is also presented.

When we stand still and do not move our eyes and head, the motion of an object in the world or the absence thereof is directly given by the motion or quiescence of the retinal image. Self-motion through the world however complicates this retinal image. During self-motion, the whole retinal image undergoes coherent global motion, called optic flow. Self-motion therefore causes the retinal motion of objects moving in the world to be confounded by a motion component due to self-motion. How then do we perceive the motion of an object in the world when we ourselves are also moving? Although non-visual information about self-motion, such as provided by efference copies of motor commands and vestibular stimulation, might play a role in this ability, it has recently been shown that the brain possesses a purely visual mechanism that underlies scene-relative object motion perception during self-motion. In the flow parsing hypothesis developed by Rushton and Warren (2005; Warren & Rushton, 2007; 2009b), the brain uses its sensitivity to optic flow to detect and globally remove retinal motion due to self-motion and recover the scene-relative motion of objects. Research into this perceptual ability has so far been of a qualitative nature. In this thesis, I therefore develop a retinal motion nulling paradigm to measure the gain with which the flow parsing mechanism uses the optic flow to remove the self-motion component from an object’s retinal motion. I use this paradigm to investigate how accurate scene-relative object motion perception during self-motion can be based on only visual information, whether this flow parsing process depends on a percept of the direction of self-motion and the tuning of flow parsing, i.e., how it is modulated by changes in various stimulus aspects. The results reveal that although adding monocular or binocular depth information to the display to precisely specify the moving object’s 3D position in the scene improved the accuracy of flow parsing, the flow parsing gain was never up to the extent required by the scene geometry. Furthermore, the flow parsing gain was lower at higher eccentricities from the focus of expansion in the flow field and was strongly modulated by changes in the motion angle between the self-motion and object motion components in the retinal motion of the moving object, the speeds of these components and the density of the flow field. Lastly, flow parsing was not affected by illusory changes in the perceived direction of self-motion. In conclusion, visual information alone is not sufficient for accurate perception of scene-relative object motion during self-motion. Furthermore, flow parsing takes the 3D position of the moving object in the scene into account and is not a uniform global subtraction process. 8e observed tuning characteristics are different from those of local perceived motion interactions, providing evidence that flow parsing is a separate process from these local motion interactions. Finally, flow parsing does not depend on a prior percept of self-motion direction and instead directly uses the input retinal motion to construct percepts of scene-relative object motion during self-motion.
published_or_final_version
Psychology
Doctoral
Doctor of Philosophy

Approved for public release; distribution is unlimited
Human performance contributes to total system performance. As human performance decreases, total system performance decreases while lifecycle costs increase. In a fiscally constrained environment, Human Systems Integration (HSI) seeks to assure human performance to reduce operating costs. This thesis seeks to develop a model for ship design in relation to Motion Induced Interruptions (MII). The model is based on the premise that MIIs affect specific domains of HSI in an adverse way. Future ship design considerations that mitigate MII occurrences can save the Navy money spent on human injury and system degradation. The thesis begins with an historical overview of MII theory and development and its interactions with domains of HSI. A MII prediction model was developed using data acquired from an experiment using a motion-based platform that emulates ship motion. Quantitative data were analyzed from 21 subjects who underwent 32 trials. Multiple regression analysis consisted of two independent variables as period and lateral acceleration and the response variable as a MII incident. Logistic regression considered two more independent variables that addressed individual differences. Data analysis revealed that acceleration, period, and human balance were statistically significant. The proposed multiple regression model accounted for 77% of the variance of MII forecasting. This thesis lays the foundation for future quantitative analysis of interactions between MIIs and accelerations or periods in different axes. Additionally, it provides an initial model that predicts conditions of high MII incident environments that can ultimately lead to a viable design tool for HSI practitioners and ship designers.

This thesis contains my writings and concept development, as it relates to personal history, time and animated art. This is explored through various methodologies, including short story and Zen meditative writing. As a companion to my recent thesis exhibition, this paper looks to make logical, emotional and spiritual connections between my art practice, ritualized cultural tropes and contemporary views of mortality.

This document explores the use of motion within design to defamiliarize a message. The objective is to expand a viewer’s level of understanding through prolonged perception. I experiment with this idea using present-day tools which afford my own movement during the capturing process to create various visual interpretations of motion. I look to László Moholy-Nagy, Aleksandr Rodchenko and Eadweard Muybridge who explored the use of the camera, the new technology of their day, to understand its potential to create a new visual language. They believed the lens of the camera was the eye of the future—and the public’s exposure to the camera’s possibilities an enlightening transcendence. I also believe in exploring newly developed tools; in visually researching their intended and unintended use to discover new perspectives in not only the way we see and represent the world, but the way we understand and represent ourselves in relation to the study of our field.

Sensitivity to differential motion components, shearing and compressive (opposed) motion, was examined. The hypothesis that the visual system contains local mechanisms specifically sensitive to these types of motion was tested. Stimuli consisted of two moving sinusoidal gratings. Sensitivity to shear and compression was compared with sensitivity for linear motion. Lower thresholds of motion and contrast sensitivities were obtained. Subjects were more sensitive to opposed than to non-opposed motion for a range of grating orientations and different grating spatial frequencies. However sensitivity for opposed motion decreased in the presence of a second added linear motion. The hypothesis of local shear and compression mechanisms was rejected in favour of antagonistic (opponent) interactions between local motion mechanisms. Motion capture was examined. Stimuli were made up of a circular test grating surrounded by another grating. Subjects were required to judge the direction of motion of the test grating. Experiments examined the effects on motion capture of: centre grating size; orientation of surround; relative contrast of centre and surround; plaids in the surround. Conditions favouring motion capture were: with the smallest centre grating; with surround and centre orientations within thirty degrees; with surround had higher contrast than the centre; and only when a plaid surround contained a component of similar orientation as the centre. For conditions of motion capture relative to those of no-capture, increased velocity thresholds for judging the centre direction were found. This was associated with a shift in the bias point between opposed directions with no change in overall sensitivity to motion. It is suggested that a cooperative network of local motion mechanisms featuring centre-surround opponency can account for all the results of this study.

Traditional musical instruments assume a high degree of mental and physical interaction, which results in the exclusion of some groups. One such group is children with severe impairments. Music is important for the development of the individuals in this group, but   also physical exercise. This thesis describes the design process of a motion-sensing MIDI controller, which is an attempt to combine music and physical exercise through the design of a new musical instrument. The prototype controller is based on an Arduino microcontroller board, a motion sensor and a Bluetooth module, and is powered by a nine-volt battery. It outputs MIDI messages automatically at regular intervals. The data is received wirelessly by a computer, whichgenerates sound based on the messages. A pilot evaluation which had a small group of test subjects play two short melodies was performed to evaluate the efficiency of the prototype. The results show that the prototype could be used to relatively quickly play through theshortest of the two melodies, while the other melody was completed relatively slowly. It is concluded that the prototype can potentiallybe used to efficiently play a shorter musical composition, but may struggle with longer ones. Future work could include conducting a larger evaluation, letting the user control when the notes are played and facilitating the configuration of user settings.

The use of conversational agents has increased across the world. From providing automated support for companies to being virtual psychologists they have moved from an academic curiosity to an application with real world relevance. While many researchers have focused on the content of the dialogue and synthetic speech to give the agents a voice, more recently animating these characters has become a topic of interest. An additional use for character animation technology is in the film and video game industry where having characters animated without needing to pay for expensive labour would save tremendous costs. When animating characters there are many aspects to consider, for example the way they walk. However, to truly assist with communication automated animation needs to duplicate the body language used when speaking. In particular conversational agents are often only an animation of the upper parts of the body, so head motion is one of the keys to a believable agent. While certain linguistic features are obvious, such as nodding to indicate agreement, research has shown that head motion also aids understanding of speech. Additionally head motion often contains emotional cues, prosodic information, and other paralinguistic information. In this thesis we will present our research into synthesising head motion using only recorded speech as input. During this research we collected a large dataset of head motion synchronised with speech, examined evaluation methodology, and developed a synthesis system. Our dataset is one of the larger ones available. From it we present some statistics about head motion in general. Including differences between read speech and story telling speech, and differences between speakers. From this we are able to draw some conclusions as to what type of source data will be the most interesting in head motion research, and if speaker-dependent models are needed for synthesis. In our examination of head motion evaluation methodology we introduce Forced Canonical Correlation Analysis (FCCA). FCCA shows the difference between head motion shaped noise and motion capture better than standard methods for objective evaluation used in the literature. We have shown that for subjective testing it is best practice to use a variation of MUltiple Stimuli with Hidden Reference and Anchor (MUSHRA) based testing, adapted for head motion. Through experimentation we have developed guidelines for the implementation of the test, and the constraints on the length. Finally we present a new system for head motion synthesis. We make use of simple templates of motion, automatically extracted from source data, that are warped to suit the speech features. Our system uses clustering to pick the small motion units, and a combined HMM and GMM based approach for determining the values of warping parameters at synthesis time. This results in highly natural looking motion that outperforms other state of the art systems. Our system requires minimal human intervention and produces believable motion. The key innovates were the new methods for segmenting head motion and creating a process similar to language modelling for synthesising head motion.

In motion pictures there is a certain amount of redundancy between consecutive frames. These redundancies can be exploited by using interframe prediction techniques. To further enhance the efficiency of interframe prediction, motion estimation and compensation, various motion compensation techniques can be used. There are two distinct techniques for motion estimation block matching and pel-recursive block matching has been widely used as it produces a better signal-to-noise ratio or a lower bit rate for transmission than the pel-recursive method. In this thesis, various pel-recursive motion estimation techniques such as steepest descent gradient algorithm have been considered and simulated.

For many years, artists and photographers have used blurred lines or `motion streaks' along an object's trajectory of motion to indicate fast motion. As it turns out, these streaks must occur in vision, because the visual system integrates information over time, around 100 - 120 ms. Generally streaks are not seen, but they could prove a useful cue to direction of motion, as suggested in an influential model proposed by Geisler (1999). In experiments exploiting the tilt aftereffect and illusion paradigms, we found that strong motion streaks produced robust tilt aftereffects and illusions, similar in magnitude and orientation tuning to those induced by tilted lines. These effects were weak or absent in weak streak conditions, and when motion was too slow to form streaks. We also investigated binocular rivalry suppression of static stimuli by fast and slow motion, and found that grating stimuli oriented parallel to the direction of fast, rivalling motions were more deeply suppressed than those orthogonal to the motion direction, but only for fast, `streaky' motion, not slow motion. We attributed this deeper suppression to within-channel masking by motion streaks, as there was clear orientation tuning of this effect, both during suppression and dominance phases. We further explored masking by motion streaks in two further studies addressing the orientation and spatial frequency tuning of dichoptic and monoptic masking by motion streaks. Finally, we used functional magnetic resonance imaging to explore the neural correlates of these streaks, and found similar patterns of activity for fast motion and static, oriented patterns, which could successfully be used by a classifier algorithm to decode whether a participant was viewing one of two directions of motion (45 or 135 degrees) after training on orientation sessions alone. Together, these results indicate that motion streaks produced by temporal integration of fast translating features effectively adapt orientation-selective cells, that they cause masking similar to that caused by static stimuli, that they can cause tuned suppression of oriented stimuli even when not seen, and that they are present in early visual cortex. Thus, motion streaks are present in the visual system, and would be available to perform the function ascribed to them by Geisler. This is discussed in terms of traditional models of motion perception, and some novel predictions and future experiments are proposed.

Positron emission tomography (PET) is a modality with high temporal resolution but long acquisition times. This can result in blurred images due to subject motion. Respiratory motion in particular is an unavoidable source of degradation, which can cause issues with quantification and clinical interpretation. An important characteristic of respiratory motion is its pseudo-cyclic nature, which has previously been exploited to form mathematical models which describe the motion, driven by a small number of parameters. The aim of this project is to use this form of motion modelling to estimate motion using information acquired from both dynamic magnetic resonance (MR) scans and from the acquired PET data itself to correct for the effects of motion. The use of motion models in this way can overcome the high levels of noise which otherwise characterise the estimation problem. First the feasibility of using motion models is investigated using synthetic data consisting of individual PET gates simulated using real motion information. The PET gates are registered using constraints provided by a motion model derived from MR images. A novelty of this approach is that this is the first time PET data have been used to indirectly drive a parameterised motion model. The next part of the project attempts to formalise the motion estimation process by incorporating the reduced-parameter motion model into the PET image reconstruction. An analytical gradient for a single motion parameter that drives the model is derived from the same objective function used to estimate the image. This results in significant noise averaging, providing robustness to the high level of noise typically found in PET data acquired over short time frames. This is shown to improve robustness to noise well enough that the number of gates can be increased and overall motion correction performance improved. Finally, the formulation is extended further to also model photon attenuation effects. This is shown to improve the performance of the algorithm when dealing with synthetic data that includes attenuation.

Textured motion - generally known as dynamic or temporal texture - is a popular research area for synthesis, segmentation and recognition. Dynamic texture is a spatially repetitive, time-varying visual pattern that forms an image sequence with certain temporal stationarity. In dynamic texture, the notion of self-similarity central to conventional image texture is extended to the spatiotemporal domain. Dynamic textures are typically videos of processes, such as waves, smoke, fire, a flag blowing in the wind, a moving escalator, or a walking crowd. Creation of synthetic frames is a key issue especially for movie screen industry to enrich their scenes from a white screen into a shining reality. In robotics world, for example an autonomous vehicle must decide what is traversable terrain (e.g. grass) and what is not (e.g. water). This problem can be addressed by classifying portions of the image into a number of categories, for instance grass, dirt, bushes or water. If these parts are identifiable, then segmentation and recognition of these textures results with an efficient path planning for the autonomous vehicle. In this thesis, we aimed to characterize these textured motions like mentioned above. We tried to implement several known techniques and compared the results.

AbstractThe wind power business has in recent times changed its focus from land-based installations to offshore installations. This has presented challenges both technological and financial, mainly related to construction and maintenance. To optimize the availability of the offshore wind turbines it is important to have support vessels and boarding systems that can handle as rough sea conditions as possible, and the relative motions between these vessels and the wind turbines become increasingly important to predict, as the offshore wind business expands.For this purpose, a need has been expressed for a simple tool for quick estimation of such motions. In this thesis, a MATLAB program has been developed for this purpose. It takes various input from the user, such as information on the sea state and the physical situation to be considered, as well as limiting criteria. The program provides the user with information on the local vessel motions and the relative motions between a point on the vessel and a fixed point on the wind turbine, and then compares it to the given criteria. It also gives out various plots to illustrate the motions and the relevant transfer functions.The final version of Relative Motion Calculator, RMC 2.3, features the following options:•Two types of wave spectra•Arbitrary placement of the moving coordinate system•Arbitrary placement of the considered points•Long- or short-crested wave theoryRMC 2.3 has undergone thorough testing to prove its validity, and all test results are reasonable and according to expectation. Although the program is a bit difficult to use, it can be used as intended, for calculating relative motions between a moving point on a vessel and a fixed point. Furthermore, the program might provide a good platform for further development.

This dissertation presents algorithms for restoring some of the major corruptions observed in archived film or video material. The two principal problems of impulsive distortion (Dirt and Sparkle or Blotches) and noise degradation are considered. There is also an algorithm for suppressing the inter-line jitter common in images decoded from noisy video signals. In the case of noise reduction and Blotch removal the thesis considers image sequences to be three dimensional signals involving evolution of features in time and space. This is necessary if any process presented is to show an improvement over standard two-dimensional techniques. It is important to recognize that consideration of image sequences must involve an appreciation of the problems incurred by the motion of objects in the scene. The most obvious implication is that due to motion, useful three dimensional processing does not necessarily proceed in a direction 'orthogonal' to the image frames. Therefore, attention is given to discussing motion estimation as it is used for image sequence processing. Some discussion is given to image sequence models and the 3D Autoregressive model is investigated. A multiresolution BM scheme is used for motion estimation throughout the major part of the thesis. Impulsive noise removal in image processing has been traditionally achieved by the use of median filter structures. A new three dimensional multilevel median structure is presented in this work with the additional use of a detector which limits the distortion caused by the filters . This technique is found to be extremely effective in practice and is an alternative to the traditional global median operation. The new median filter is shown to be superior to those previously presented with respect to the ability to reject the kind of distortion found in practice. A model based technique using the 3D AR model is also developed for detecting and removing Blotches. This technique achieves better fidelity at the expense of heavier computational load. Motion compensated 3D IIR and FIR Wiener filters are investigated with respect to their ability to reject noise in an image sequence. They are compared to several algorithms previously presented which are purely temporal in nature. The filters presented are found to be effective and compare favourably to the other algorithms. The 3D filtering process is superior to the purely temporal process as expected. The algorithm that is presented for suppressing inter-line jitter uses a 2D AR model to estimate and correct the relative displacements between the lines. The output image is much more satisfactory to the observer although in a severe case some drift of image features is to be expected. A suggestion for removing this drift is presented in the conclusions. There are several remaining problems in moving video. In particular, line scratches and picture shake/roll. Line scratches cannot be detected successfully by the detectors presented and so cannot be removed efficiently. Suppressing shake and roll involves compensating the entire frame for motion and there is a need to separate global from local motion. These difficulties provide ample opportunity for further research.

The theory of Brownian motion is a cornerstone of modern physics. In this thesis, we introduce a nonequilibrium extension to this theory, namely an effective Markovian theory of the Brownian motion of a heated nanoparticle. This phenomenon belongs to the class of nonequilibrium steady states (NESS) and is characterized by spatially inhomogeneous temperature and viscosity fields extending in the solvent surrounding the nanoparticle. The first chapter provides a pedagogic introduction to the subject and a concise summary of our main results and summarizes their implications for future developments and innovative applications. The derivation of our main results is based on the theory of fluctuating hydrodynamics, which we introduce and extend to NESS conditions, in the second chapter. We derive the effective temperature and the effective friction coefficient for the generalized Langevin equation describing the Brownian motion of a heated nanoparticle. As major results, we find that these parameters obey a generalized Stokes–Einstein relation, and that, to first order in the temperature increment of the particle, the effective temperature is given in terms of a set of universal numbers. In chapters three and four, these basic results are made explicit for various realizations of hot Brownian motion. We show in detail, that different degrees of freedom are governed by distinct effective parameters, and we calculate these for the rotational and translational motion of heated nanobeads and nanorods. Whenever possible, analytic results are provided, and numerically accurate approximation methods are devised otherwise. To test and validate all our theoretical predictions, we present large-scale molecular dynamics simulations of a Lennard-Jones system, in chapter five. These implement a state-of-the-art GPU-powered parallel algorithm, contributed by D. Chakraborty. Further support for our theory comes from recent experimental observations of gold nanobeads and nanorods made in the the groups of F. Cichos and M. Orrit. We introduce the theoretical concept of PhoCS, an innovative technique which puts the selective heating of nanoscopic tracer particles to good use. We conclude in chapter six with some preliminary results about the self-phoretic motion of so-called Janus particles. These two-faced hybrids with a hotter and a cooler side perform a persistent random walk with the persistence only limited by their hot rotational Brownian motion. Such particles could act as versatile laser-controlled nanotransporters or nanomachines, to mention just the most obvious future nanotechnological applications of hot Brownian motion.

Hjärnskador är vanligt förekommande inom ishockey och för att kunna utveckla bättre skyddsutrustning krävs information om kroppens rörelse. Med hjälp av motion tracking kan man få fram rörelsers och kollisioners hastigheter. Automatiserad motion tracking skulle kunna utesluta felkällor till följd av användare etc.             Två olika Matlabprogram för motion tracking rekonstruerades för att undersöka möjligheten att i framtiden automatisera tekniken. Leonid Sigal, forskare vid Disney research Pittsburgh, har utvecklat ett program som gick att rekonstruera och använda på utvalda videosekvenser. För att uppnå korrekt motion tracking måste initialiseringen för varje videosekvens beräknas.
Brain injuries are common in ice hockey and knowledge about the body and its movement is necessary to develop safety equipment to prevent such injuries. The velocity of the movement in collisions can be detected with motion tracking. Automatic motion tracking can reduce the impact of error sources such as those caused by the user. Two different Matlab programs were reconstructed to analyse the possibilities to develop an automatic technique for motion tracking. Leonid Sigal, scientist at Disney research Pittsburgh, has developed one of the programs for motion tracking. It was possible to reconstruct his program and use it for selected video sequences. Initial values for each video sequence had to be calculated to achieve accurate motion tracking.

Mobile robots are increasingly being deployed in the real world in response to a heightened demand for applications such as transportation, delivery and inspection. The motion planning systems for these robots are expected to have consistent performance across the wide range of scenarios that they encounter. While state-of-the-art planners, with provable worst-case guarantees, can be employed to solve these planning problems, their finite time performance varies across scenarios. This thesis proposes that the planning module for a robot must adapt its search strategy to the distribution of planning problems encountered to achieve real-time performance. We address three principal challenges of this problem. Firstly, we show that even when the planning problem distribution is fixed, designing a nonadaptive planner can be challenging as the performance of planning strategies fluctuates with small changes in the environment. We characterize the existence of complementary strategies and propose to hedge our bets by executing a diverse ensemble of planners. Secondly, when the distribution is varying, we require a meta-planner that can automatically select such an ensemble from a library of black-box planners. We show that greedily training a list of predictors to focus on failure cases leads to an effective meta-planner. For situations where we have no training data, we show that we can learn an ensemble on-the-fly by adopting algorithms from online paging theory. Thirdly, in the interest of efficiency, we require a white-box planner that directly adapts its search strategy during a planning cycle. We propose an efficient procedure for training adaptive search heuristics in a data-driven imitation learning framework. We also draw a novel connection to Bayesian active learning, and propose algorithms to adaptively evaluate edges of a graph. Our approach leads to the synthesis of a robust real-time planning module that allows a UAV to navigate seamlessly across environments and speed-regimes. We evaluate our framework on a spectrum of planning problems and show closed-loop results on 3 UAV platforms - a full-scale autonomous helicopter, a large scale hexarotor and a small quadrotor. While the thesis was motivated by mobile robots, we have shown that the individual algorithms are broadly applicable to other problem domains such as informative path planning and manipulation planning.We also establish novel connections between the disparate fields of motion planning and active learning, imitation learning and online paging which opens doors to several new research problems.

In the development of multicellular organisms, tissue development and homeostasis require coordinated cellular motion. For example, in conditions such as wound healing, immune and epithelial cells need to proliferate and migrate. Deregulation of key signalling pathways in pathological conditions causes alterations in cellular motion properties that are critical for disease development and progression, in cancer it leads to invasion and metastasis. Consequently there is strong interest in identifying factors, including drugs that affect the motion and interactions of cells in disease using experimental models suitable for high-content screening. There are two main modes of cell migration; individual and collective migration. Currently analysis tools for robust, sensitive and comprehensive motion characterisation in varying experimental conditions for large extended timelapse acquisitions that jointly considers both modes are limited. We have developed a systematic motion analysis framework, Motion Sensing Superpixels (MOSES) to quantitatively capture cellular motion in timelapse microscopy videos suitable for high-content screening. MOSES builds upon established computer vision approaches to deliver a minimal parameter, robust algorithm that can i) extract reliable phenomena-relevant motion metrics, ii) discover spatiotemporal salient motion patterns and iii) facilitate unbiased analysis with little prior knowledge through unique motion 'signatures'. The framework was validated by application to numerous datasets including YouTube videos, zebrafish immunosurveillance and Drosophila embryo development. We demonstrate two extended applications; the analysis of interactions between two epithelial populations in 2D culture using cell lines of the squamous and columnar epithelia from human normal esophagus, Barrett's esophagus and esophageal adenocarcinoma and the automatic monitoring of 3D organoid culture growth captured through label-free phase contrast microscopy. MOSES found unique boundary formation between squamous and columnar cells and could measure subtle changes in boundary formation due to external stimuli. MOSES automatically segments the motion and shape of multiple organoids even if present in the same field of view. Automated analysis of intestinal organoid branching following treatment agrees with independent RNA-seq results.

We are taught history abstractly as a time-line; sequentially. In reality, history presents itself as an accumulation of layers of happenings and occurrences. These layers are present in the built environment of the city as it is today. Similarly, the layers of culture are present in the daily rituals and language of the present. In the case of a train station, layers present themselves as layers of velocity. It is the place where pedestrians, automobiles and trains converge and individuals move between varying degrees of motion. The challenge in the built world today is to make a place for all layers of motion, but in a hierarchical fashion so that the individual (the pedestrian) is honored. If the pedestrian is relegated a place below machines which facilitate movement, then the city suffers. "In the narrow streets of the medieval town, persuasion is mainly through the sight and smell of real cakes in the window. On the commercial strip the supermarket windows contain no merchandise. [there are signs instead]" - Robert Venturi Learning from Las Vegas When the automobile is given the place of honor, the sign is more important than the experience. The speed at the motorist scale does not allow for any experience other than symbol or sign. This reduces the sensual experience of the individual's life and results in the collective loss of place and rootedness. Just as the individual's memory can be triggered by smells, sounds or textures, the city's memory is also made of these things. Americans seem to be constantly in search of roots. The majority of objects and places in American cities seems to cater to the motorist scale and for this reason there can be no roots. At best there is only a symbol for roots, like a coat of arms with no meaning.
Master of Architecture

The goal of this thesis is to contribute to the ambitious program of the foundation of developing statistical physics using chaos. We build a deterministic model of Brownian motion and provide a microscpoic derivation of the Fokker-Planck equation. Since the Brownian motion of a particle is the result of the competing processes of diffusion and dissipation, we create a model where both diffusion and dissipation originate from the same deterministic mechanism - the deterministic interaction of that particle with its environment. We show that standard diffusion which is the basis of the Fokker-Planck equation rests on the Central Limit Theorem, and, consequently, on the possibility of deriving it from a deterministic process with a quickly decaying correlation function. The sensitive dependence on initial conditions, one of the defining properties of chaos insures this rapid decay. We carefully address the problem of deriving dissipation from the interaction of a particle with a fully deterministic nonlinear bath, that we term the booster. We show that the solution of this problem essentially rests on the linear response of a booster to an external perturbation. This raises a long-standing problem concerned with Kubo's Linear Response Theory and the strong criticism against it by van Kampen. Kubo's theory is based on a perturbation treatment of the Liouville equation, which, in turn, is expected to be totally equivalent to a first-order perturbation treatment of single trajectories. Since the boosters are chaotic, and chaos is essential to generate diffusion, the single trajectories are highly unstable and do not respond linearly to weak external perturbation. We adopt chaotic maps as boosters of a Brownian particle, and therefore address the problem of the response of a chaotic booster to an external perturbation. We notice that a fully chaotic map is characterized by an invariant measure which is a continuous function of the control parameters of the map. Consequently if the external perturbation is made to act on a control parameter of the map, we show that the booster distribution undergoes slight modifications as an effect of the weak external perturbation, thereby leading to a linear response of the mean value of the perturbed variable of the booster. This approach to linear response completely bypasses the criticism of van Kampen. The joint use of these two phenomena, diffusion and friction stemming from the interaction of the Brownian particle with the same booster, makes the microscopic derivation of a Fokker-Planck equation and Brownian motion, possible.

Bakgrund: Depression förutspås bli den näst största sjukdomsbördan i världen år 2020. Idag återfinns psykoterapi och läkemedel som behandlingsmetoder mot depressioner. Utöver detta har studier visat på ett samband mellan minskade depressiva symtom och fysisk aktivitet. Metod: Material för granskning har sammanställts genom sökningar i artikeldatabaserna PubMed, CINAHL, PsycINFO och granskats enligt riktlinjer av Statens beredning för medicinsk utvärdering (SBU). Genom en manifest innehållsanalys har bärande meningar identifierats och lagt grunden för resultatet. Syfte: Att med en litteraturstudie undersöka vilken inverkan fysisk aktivitet har på depressionstillstånd, samt att undersöka vilka omvårdnads-åtgärder relaterat till fysiska aktiviteter den legitimerade sjuksköterskan kan rekommendera för att främja hälsan hos patienten. Resultat: Fysisk aktivitet ger minskade symtom samt förebygger insjuknande i depression. Mängden och intensiteten har ett positivt samband med minskade depressiva symtom. Slutsats: Fysisk aktivitet ger en hälsovinst för patienter med depressiva symtom. Mer forskning krävs för att uppnå en vetenskaplig trygghet i behandlingsformen.
Background: Depression is predicted to become the second largest disease burden in the world in 2020. Present treatment options for depression include psychotherapy and pharmacotherapy. In addition, studies have linked increased physical activity with decreased depressive symptoms. Method: Material for review was compiled by searching the article databases PubMed , CINAHL , PsycINFO and reviewed in accordance to the guidelines provided by the Swedish Council on Health Technology Assessment (SBU). Meaningful sentences has been identified through a manifest content analysis and served as the foundation for the results. Aim: To, as a review, examine the impact of physical activity on depression and to examine how the. Results: Physical activity reduced symptoms and prevented the onset of depression. The amount and intensity has a positive correlation with decreased depressive symptoms. Conclusion: Physical activity provides a health benefit for patients with depressive symptoms. More research is needed to scientifically confirm the treatment form.

Set in Motion is a composition in two movements for an ensemble of 10 instruments that interacts with a computer in real-time performance using Max/MSP software. The instrumental writing in the piece focuses on incorporating electronic composition techniques into acoustic composition. Techniques such as constructing envelopes, shading, and sound masses, which are common in electronic works are applied to the ensemble. Also, the intervals of a major and minor second are important in the instrumental writing. These intervals provided the basis for the material, which was then transformed algorithmically, both in individual lines and whole sections of music. In terms of the electronics, the first movement takes advantage of the interactive possibilities of the software. The electronic part is created through the computer extending, processing, and responding to the instruments. The techniques include extracting small grains of sound from the live instruments and repeating them, analyzing the pitch of the signal and amplifying overtones, and rearranging fragments in blocks of recorded sound. The second movement changes the focus of the electronic sound. The electronics are made of precomposed gestures that compliment the live instruments. These gestures are categorized and selected randomly according to specific characteristics. Filtering and other effects applied to the sounds are chosen randomly as well.

In the last few years a lot of works have been done in the field of motor control and in motion analysis. Several different hypotheses have been described and reviewed to understand living beings on motor coordination. What is commonly referred to, as motor control is indeed an articulated problem that is, at least from a robotic perspective, often more suitably divided in: sensing (perception, cognition), deliberation, planning, kinematic control and dynamic control. Through the pages of this work, several different problems related to motion control, to living beings motion and to its robotic counterpart will be addressed. The strong underling motif of all the proposed algorithms and architectures (both software and hardware) is the presence of a real environment interaction. From reflexes to motion planning, from architecture definition to smart sensor design and from kinematic modelling of the human body to the action-perception loop implementation, this thesis intends to be a first gatherer of information and ideas in this interesting and complex field.

Human motion analysis in the video has its vast application. The recognition of the human action is the most prominent application of human motion analysis. In this research we analyzed different methodologies for modeling human action. We also discussed challenges and methodologies which are used to handle them. These methodologies are divided into two categories. One is global feature descriptor and other is local feature descriptors. The disadvantage of the global feature descriptor is that they can only give the shape information but fails to give motion information. The local feature descriptors are used to find out the motion information of the action video. The disadvantage is that it cannot give the shape or structure information of the action video. The hybrid descriptors are used to solve these problems but these descriptors suffer from high dimensionality features. In this research we proposed new feature descriptors which are capable to deal with these issues in the following manner: 1) We proposed a new local descriptor evaluated from the Finite Element Analysis for human action recognition. This local descriptor represents the distinctive human poses in the form of the stiffness matrix. This stiffness matrix gives the information of motion as well as shape change of the human body while performing an action. Initially, the human body is represented in the silhouette form. Most prominent points of the silhouette are then selected. This silhouette is discretized into several finite small triangle faces (elements) where the prominent points of the boundaries are the vertices of the triangles. The stiffness matrix of each triangle is then calculated. The feature vector representing the action video iii frame is constructed by combining all stiffness matrices of all possible triangles. These feature vectors are given to the Radial Basis Function-Support Vector Machine (RBF-SVM) classifier. The proposed method shows its superiority over other existing state-of-the-art methods on the challenging datasets Weizmann, KTH, Ballet, and IXMAS. 2) Background cluttering, appearance change due to variation in viewpoint and occlusion are the prominent hurdles that can reduce the recognition rate significantly. Methodologies based on Bag-of-visual-words are very popular because they do not require accurate background subtraction. But the main disadvantage with these methods is that they do not retain the geometrical structural information of the clusters that they form. As a result, they show intra-class mismatching. Furthermore, these methods are very sensitive to noise. Addition of noise in the cluster also results in the misclassification of the action. To overcome these problems we proposed a new approach based on modified Bag-of-visual-word. Proposed methodology retains the geometrical structural information of the cluster based on the calculation of contextual distance among the points of the cluster. Normally contextual distance based on Euclidean measure cannot deal with the noise but in the proposed methodology contextual distance is calculated on the basis of a difference between the contributions of cluster points to maintain its geometrical structure. Later directed graphs of all clusters are formed and these directed graphs are described by the Laplacian. Then the feature vectors representing Laplacian are fed to the Radial Basis Function based Support Vector Machine (RBF-SVM) classifier. iv 3) We also proposed a new feature descriptor to detect abnormality in a video captured for surveillance applications in real-time and also overcome the problem of the curse of dimensionality. To extract features related to any change in the video, non linear Gaussian fuzzy lattice functions have been applied on each frame of the video which results in the formation of fuzzy lattices. These fuzzy lattices have been expressed in the form of Schrödinger equation to find the kinetic energy involved corresponding to any change in the video. A number of the fuzzy lattice has been used as a dimension of the feature. It reduces the dimensionality significantly as compared to other state-of-the-art methods. Finally, the kinetic energy parameter is classified into normal and abnormal activities with the help of SVM-based classifier.

A striking number of psychophysical, neuropsychological, and human neuroimaging studies suggest the presence of distinct mechanisms and neural substrates for the perception of motion defined by spatiotemporal variations of luminance (first-order motion) and the perception of motion defined by spatiotemporal variations of contrast (second-order motion) (Ledgeway & Smith, 1994a,b; Lu & Sperling, 1995, 2001). The experiments outlined in this Thesis aimed to investigate further the mechanisms underlying the perception of first- and second-order motion and how these two motion signals are integrated by the visual system. To this end, first-order, second-order and cross-order motion conditions were tested under different experimental paradigms such as motion priming (Chapter 2), motion aftereffect (MAE - Chapter 3) and motion-induced position shift (MIPS - Chapter 4). In particular, the experiments were designed to possibly tap low-levels, intermediate-levels and high-levels of motion processing, to understand better whether first- and second-order motion perception involve different and separate mechanisms from the lower level of motion analysis (e.g., V1, V2/V3) and whether these two kinds of motion signals are integrated at a higher level of motion analysis. In the first experiment (Chapter 5) we investigated the implicit short-term memory mechanisms for first- and second-order moving stimuli. In particular, we used a repetition priming paradigm (Chapter 2) in order to test if priming for motion direction is sensitive to spatial position with both first- and second-order motion. We also used a cross-order condition in which first-order motion patterns primed second-order moving patterns and vice versa. Testing priming for motion direction and spatial position separately for first- and second-order motion allowed us to assess if these two motion cues are represented and stored in neural substrates with a retinotopical organization. The cross-order condition permitted us to assess whether the implicit short-term memory for first- and second-order motion relies on the same mechanism/s and neural substrates; that is, if first-order primes second-order motion and vice versa, this would suggest a common locus of priming representation for first- and second-order motion. We found that priming for motion direction occurs both with first- and second-order motion. Moreover, priming for motion direction is position-sensitive both with first- and second-order motion, suggesting a neural locus of priming representation with a retinotopical organization. Previous findings of Campana et al. (2002, 2006) have shown the involvement of cortical area MT in priming for motion direction using first-order stimuli, based on the results of the first experiment we suggest that MT might be involved also in second-order motion priming. We found that cross-order motion priming also exists but it is weaker and not sensitive to spatial position. From these findings we hypothesize that first- and second-order motion cues remain distinct and separate at the level in which global motion is extracted (i.e., MT; Edwards & Badcock, 1995). However, the two types of motion could be integrated at a higher level of motion processing, where the retinotopic organization is either lost or at least very coarse (e.g., MST). The findings of the first experiment showed both that, in the within-order conditions priming for motion direction is dependent on the repetition of the same target’s position; and, in the cross-order condition, priming is not sensitive to spatial position. These findings suggest not only there are two independent pathways for first- and second-order motion, but also that these two motion processing streams seem to encode the spatial position of a moving pattern separately. In the second experiment (Chapter 6) we assessed more specifically the issue of whether first- and second-order motion encode and assign the position of a moving pattern by means of a single and common mechanism, or whether there are two distinct mechanisms for position assignment. To this purpose we measured the motion-induced position shift (MIPS) for first-order and second-order drifting Gabors (De Valois & De Valois, 1991; Durant & Johnston, 2004; Edwards & Badcock, 2003; Fang & He, 2004). We first measured the MIPS for first- and second-order moving patterns separately, then conducted another experiment in which first- and second-order drifting Gabors were presented within the same trial, to see if cross-order motion shifts perceived position as well. If there is a common position assignment mechanism for both first- and second-order motion cues, one would expect an effect size of cross-order MIPS intermediate between that obtained with first- and second-order motion separately. On the other hand, the lack of an effect for cross-order stimuli would indicate the presence of independent position assignment mechanisms. We found that both first- and second-order motion, presented separately, shift the perceived position. We did not find any positional shift with cross-order stimuli. This implies the presence of separate mechanisms that encode and assign the spatial position for these two motion cues. In the third experiment (Chapter 7), we used a paradigm developed by Kanai and Verstraten (2005) in order to tap the mechanisms of motion processing (of both first- and second-order motion) and their interaction with adaptation at low- (e.g., V1) and intermediate levels (e.g., MT) of motion processing. Kanai and Verstraten (2005) showed that depending both on the duration of the adapting stimulus and that of the adaptation-test blank interval (i.e., inter-stimulus intervals [ISI]), the perceived direction of an ambiguous test pattern can be biased towards the opposite direction of the adaptation pattern, or in the same direction. Specifically, at very short ISIs (below 120 ms), very short adaptation durations (below 100 ms) can produce a rapid form of priming (rapid visual motion priming [rVMP]), whereas slightly longer adaptation durations (e.g., 320 and 640 ms) can yield a rapid form of motion aftereffect (rapid motion aftereffect [rMAE]). Moreover, Kanai and Verstraten (2005) showed that using even longer adaptation durations and ISI longer than one second biases the perceived motion direction of the test pattern toward the same motion direction (i.e., motion priming), obtaining a kind of priming for motion direction similar to that obtained in our first experiment. This probably reflects the activity of intermediate-level areas (e.g., MT). Very brief adaptation durations could selectively tap the response of low-level first- and second-order motion detectors. In this experiment we assessed if rVMP, rMAE, and the motion priming obtained with longer ISI also exist within the second-order motion domain. We found that the rapid effects (i.e., rVMP and rMAE) and motion priming had similar time courses for first- and second-order motions when presented separately. In a cross-order adaptation condition (i.e., adapting to first-order then testing with second-order, and vice versa) we found asymmetric transfers between first- and second-order motion cues. First-order motion influenced the processing of second-order motion but not vice versa (resulting only in rMAEs) (see Schofield et al., 2007). Taken together, our findings support the notion that first- and second-order motion are encoded by separate mechanisms from the early stages of motion processing but could be integrated at higher levels of motion processing (as in the case of the cross-order priming condition). However, the last experiment showed a certain degree of asymmetric transfer between first- and second-order motions: indeed, we found rMAEs only when we adapted to first-order and tested with second-order. These results imply that, even in low-level processing, first-order motion can influence the perception of second-order motion but not vice versa, further implying a hierarchical organization to the early mechanisms responding to these two motion cues.
Numerosi studi psicofisici, neuropsicologici e di neuroimmagine hanno evidenziato la presenza di meccanismi e substrati neurali distinti per la percezione del movimento di primo ordine (definito da variazioni spazio-temporali di luminanza) e di secondo ordine (definito, ad esempio, da variazioni spazio-temporali di contrasto) (Ledgeway & Smith, 1994a,b; Lu & Sperling, 1995, 2001). Gli esperimenti condotti in questa tesi hanno avuto l’obiettivo di studiare e chiarire ulteriormente i meccanismi sottostanti la percezione del movimento di primo e secondo ordine, e come questi due tipi di segnale sono integrati dal sistema visivo. A questo proposito è stato testato movimento di primo ordine, di secondo ordine e cross ordine utilizzando differenti paradigmi sperimentali, come il motion priming (Capitolo 2), il motion aftereffect (MAE – Capitolo 3) e il motion-induced position shift (spostamento indotto dal movimento della posizione percepita [MIPS] – Capitolo 4). In particolare gli esperimenti condotti sono stati progettati per elicitare attività neurale a livelli bassi, intermedi ed alti d’elaborazione del movimento. I suddetti esperimenti sono stati condotti per comprendere se la percezione del movimento di primo e secondo ordine richiede differenti e distinti meccanismi dai livelli più bassi di analisi (come ad esempio V1, V2/V3), e se questi due tipi di movimento sono o meno integrati ad un più alto livello di elaborazione del movimento. Nel primo esperimento (Capitolo 5) sono stati investigati i meccanismi di memoria implicita a breve termine per entrambi i tipi di movimento utilizzando il paradigma del repetition priming (Capitolo 2). In particolare è stato testato se il priming per la direzione di movimento dipende o meno dalla posizione spaziale del target. Inoltre è stata utilizzata anche una condizione cross-ordine nella quale uno stimolo di primo ordine fungeva da prime ad uno stimolo di secondo ordine, e viceversa. Testare la presenza di un effetto priming per la direzione di movimento e posizione spaziale per movimento di primo e di secondo ordine, ha permesso di studiare se questi segnali di movimento sono rappresentati in un’area corticale organizzata in modo retinotopico. La condizione cross-ordine invece ha permesso di investigare se la memoria implicita a breve termine per primo e secondo ordine dipende da uno stesso meccanismo/i e da una stessa area/e corticale/i. Se uno stimolo di primo ordine facilitasse la risposta ad uno stimolo di secondo ordine e viceversa, questo indicherebbe il coinvolgimento della stessa area corticale nella rappresentazione del priming (cioè della traccia mnestica relativa alla direzione di movimento). I risultati hanno mostrato che l’effetto di priming per la direzione di movimento avviene sia per stimoli di primo che di secondo ordine. Inoltre quest’effetto è dipendente dalla posizione spaziale del target per entrambi i tipi di movimento. Questo risultato indica chiaramente il coinvolgimento di un area/e corticale/i con organizzazione retinotopica. Precedenti studi hanno dimostrato il coinvolgimento dell’area medio-temporale (MT) nel priming per la direzione di movimento, ma ciò è stato messo in evidenza solo per stimoli di primo ordine (Campana et al., 2002, 2006). Considerando questi studi è possibile supporre il coinvolgimento della medesima area anche per il priming per la direzione di movimento con stimoli di secondo ordine. Inoltre, è stato ottenuto priming per la direzione di movimento anche nella condizione cross-ordine, anche se l’effetto è molto debole se confrontato con quello ottenuto nelle condizioni in cui primo e secondo ordine sono presentati separatamente. Tuttavia, il priming cross-ordine non è dipendente dalla posizione spaziale del target. Questi risultati supportano l’ipotesi della presenza di meccanismi indipendenti che rispondono a movimento di primo e secondo ordine. Questi meccanismi potrebbero mantenersi separati fino al livello in cui è estratto il movimento globale (es. MT) (Edwards & Badcock, 1995). Tuttavia i segnali di movimento di primo e di secondo ordine potrebbero essere integrati ad un più alto livello di analisi, dove viene meno l’organizzazione retinotopica, come ad esempio nell’area medio-temporale superiore (MST). I risultati del primo esperimento hanno dimostrato come, nelle condizioni in cui primo e secondo ordine sono testati separatamente, l’effetto di priming per la direzione di movimento risulti essere dipendente dalla posizione spaziale del target, mentre questa dipendenza dalla posizione spaziale viene meno nella condizione cross-ordine. Questi risultati non solo stanno ad indicare la presenza di due meccanismi separati per primo e secondo ordine, ma anche che questi meccanismi sono in grado di codificare ed assegnare indipendentemente la posizione spaziale di uno stimolo in movimento. Nel secondo esperimento (Capitolo 6) è stata testata l’ipotesi relativa alla presenza di un unico meccanismo responsabile della codifica e dell’assegnamento della posizione spaziale per stimoli di primo e secondo ordine, oppure di due meccanismi separati deputati a questo tipo di operazioni. A questo fine è stata misurata l’entità dello spostamento percepito della posizione spaziale indotto da stimoli in movimento (MIPS) (De Valois & De Valois, 1991; Durant & Johnston, 2004; Edwards & Badcock, 2003). In primo luogo è stato misurato esclusivamente lo spostamento della posizione spaziale percepita indotto dalla presentazione di soli stimoli di primo ordine o di soli stimoli di secondo ordine. In un secondo momento è stata utilizzata una condizione in cui gli stimoli di primo e di secondo ordine apparivano all’interno della stessa prova, per verificare se il movimento cross-ordine potesse influenzare o meno la posizione percepita degli stimoli. La logica sottostante a questa condizione è la seguente: nel caso di un meccanismo comune di codifica ed assegnazione della posizione spaziale sia per stimoli di primo che di secondo ordine, ci si aspetterebbe un effetto intermedio a quello ottenuto presentando separatamente questi due stimoli. D’altra parte, la mancanza di un effetto nella condizione cross-ordine indicherebbe la presenza di meccanismi separati ed indipendenti nella codifica e l’assegnazione della posizione spaziale per stimoli di primo e di secondo ordine. I risultati mostrano che sia il movimento di primo che di secondo ordine, quando presentati separatamente, sono in grado d’influenzare la percezione della posizione spaziale di uno stimolo in movimento. Tuttavia non è stato trovato nessun effetto nella condizione cross-ordine. Questo implica chiaramente la presenza di distinti meccanismi che codificano ed assegnano la posizione spaziale per stimoli di movimento di primo e di secondo ordine. I risultati dei primi due esperimenti supportano l’esistenza di meccanismi distinti per la percezione di movimento di primo e secondo ordine. I dati presentati in questa tesi in accordo con numerosi studi psicofisici, hanno messo in evidenza che i meccanismi che rispondono al movimento di primo e di secondo ordine sono sensibili a differenti frequenze spaziali e temporali. Tuttavia gli studi considerati finora hanno sempre utilizzato lunghi periodi di esposizione degli stimoli, è possibile pertanto che la maggior parte dei risultati ottenuti siano stati influenzati da meccanismi attentivi. Recentemente è stato dimostrato come esposizioni molto brevi (es. 80 ms) a stimoli direzionali di primo ordine sono in grado di far percepire come unidirezionale uno stimolo ambiguo (cioè uno stimolo per il quale non è possibile distinguere nettamente un movimento verso destra o verso sinistra) presentato successivamente (Kanai & Verstraten, 2005). L’utilizzo di questo paradigma può essere utile per lo studio dell’attività di detettori di movimento di basso livello che rispondono al movimento di primo e di secondo ordine. Il paradigma inoltre potrebbe servire per studiare se i meccanismi che rispondono al movimento di primo e di secondo ordine sono separati sin dai livelli più bassi di analisi del movimento (es. V1, V2/V3). E’stato dimostrato che, dipendentemente dalla durata dell’adattamento e dalla durata dell’intervallo tra la presentazione dello stimolo di adattamento e quello test (Intervallo-Inter-Stimolo, IIS), lo stimolo test ambiguo può essere percepito in movimento nella direzione contraria a quella del pattern di adattamento (motion aftereffect rapido - rMAE) oppure nella stessa direzione rispetto allo stimolo di adattamento (visual motion priming rapido - rVMP). Tuttavia Kanai e Verstraten (2005) hanno dimostrato che, utilizzando periodi di adattamento di circa 300 ms e IIS più lunghi di 2 secondi lo stimolo test ambiguo è percepito in movimento nuovamente nella medesima direzione del pattern di adattamento. Gli autori hanno chiamato questo effetto “Sensibilizzazione Percettiva” (SP), un fenomeno molto simile all’effetto di priming descritto nel primo esperimento. Questo effetto potrebbe indicare attività a livelli più alti d’elaborazione del movimento come ad esempio MT. L’effetto di sensibilizzazione percettiva emerge gradualmente nel tempo. Nel terzo esperimento (Capitolo 7) gli effetti rVMP, rMAE e SP sono stati studiati sia nel dominio del movimento di primo ordine che in quello di secondo ordine. Anche in questo caso è stata utilizzata una condizione cross-ordine in cui si è adattato con movimento di primo ordine e si è tesato con movimento di secondo ordine e viceversa. I risultati hanno mostrato che gli effetti rMAE, rVMP e SP sono molto simili nelle condizioni in cui gli stimoli di primo e di secondo ordine sono stati presentati separatamente. I risultati della condizione cross-ordine mostrano un trasferimento asimmetrico dell’adattamento, in particolare sono stati ottenuti esclusivamente rMAE adattando a primo ordine e testando con secondo ordine. Non è emerso nessun effetto adattando a movimento di secondo ordine e testando con movimento di primo ordine (Schofield et al., 2007). I risultati degli esperimenti presentati in questa tesi indicano che, a parte la presenza di decorsi temporali simili per primo e secondo ordine, i meccanismi sottostanti la percezione di questi due tipi di movimento sembrano essere separati dai più bassi livelli di analisi del movimento fino ad includere il livello in cui avviene l’estrazione del movimento globale (MT) (Edwards & Badcock, 1995; Campana et al., 2008). Nel terzo esperimento è emerso, infatti, che gli effetti rapidi di MAE e VMP si trasferiscono in modo asimmetrico quando gli stimoli di adattamento e test sono di diverso tipo, suggerendo un’organizzazione gerarchica dei meccanismi che rispondono a movimento di primo e secondo ordine. In particolare sembra che il meccanismo che risponde a movimento di primo ordine sia precoce e in grado d’influenzare l’elaborazione del movimento anche a più alti livelli di analisi, e sembra influenzi anche l’elaborazione del movimento di secondo ordine. Tuttavia il meccanismo che risponde al movimento di secondo ordine sembra non influenzare l’elaborazione del movimento di primo ordine. Questa organizzazione gerarchica potrebbe spiegare gli effetti asimmetrici riscontrati nel terzo esperimento. I risultati delineati in questa tesi supportano l’ipotesi che questi due tipi di movimento siano elaborati da differenti meccanismi e differenti popolazioni neurali, probabilmente presenti al livello delle stesse aree corticali: questo potrebbe essere plausibile sia per bassi, intermedi che a più elevati livelli di analisi del movimento. In base ai dati presentati è possibile che questi due segnali di movimento siano integrati ad un alto livello di analisi, come ad esempio MST. Tuttavia sono necessari altri studi per meglio capire quale sia l’area o le aree corticali in cui i segnali di movimento di primo e secondo ordine sono integrati.

The focus of this thesis is the tracking of apparent motion from sequences of meteorological satellite images, in the context of Atmospheric Motion Vector (AMV) derivation. AMVs are estimates of atmospheric wind, they are routinely produced from visible, infrared window and water vapour (WV) imagery, and they represent a major contribution to the observation of the Earth's atmosphere. This research tackles a number of issues related to the derivation of AMVs from WV imagery, using region-matching methods for tracking motion. WV imagery is particularly challenging, as images typically have a soft appearance, with no edges, no background, and large areas of low contrast. The datasets used in the experiments are real sequences of images in the WV 6.2 urn band from the geostationary satellite Meteosat-9. The underlying approach throughout the thesis is Gaussian multi-scale representation, a sound mathematical framework, developed by the computer vision community, that allows to analyse images at differente scales and to handle partial derivatives in a way deeply connected to the scale. The main contributions of this thesis are: It shows how Gaussian multi-scale representation can be used with WV meteorological satellite imagery, and in particular its value to handle scales and spatial derivatives. It proposes a new method to detect locations of interest, based on a difference of Gaussians, and shows that it performs better than other detectors, including those commonly used in operational AMV derivation schemes, in the experiments carried out with Meteosat-9 imagery. It proposes a group of distances, based on the Sobolev norm HI, which includes a term to evaluate partial derivatives, to quantify the similarity between neighbourhoods, and it shows that these distances produce better results than the widely used £2 norm, especially when the weight given to the derivatives term is relatively large.

Throughout the current development of video coding technologies, the main improvements are increasing the number of possible prediction directions and adding more sizes and more modes for blocks coding. However, there are no major substantial changes in video coding technology. The conventional video coding algorithms works well for video with motions of directions parallel to the image plane, but their efficiency drops for other kinds of motions, such as dolly motions. But increasing number of videos are captured by moving cameras as the video devices are becoming more diversified and lighter. Therefore, a higher efficient video coding tool has to be used to compress the video for new video technologies. In this thesis, a novel video coding tool, Global Motion Estimation using Motion Sensor (GMEMS), is proposed. Then, a series related approaches are researched and evaluated. The main target of this tool is using advanced motion sensor technology and computer graphics tools to improve and extend the traditional motion estimation and compensation method, which could finally enhance the video coding efficiency. Meanwhile, the computational complexity of motion estimation method is reduced as some differences have been compensated. Firstly, a Motion information based Coding method for Texture sequences (MCT) is proposed and evaluated using H.264/AVC standard. In this method, a motion sensor commonly-used in smart-phones is employed to get the panning motion (rotational motion). The proposed method could compensate panning motion by using frame projection using camera motion and a new reference allocation method. The experimental results demonstrate the average video coding gain is around 0.3 dB. In order to apply this method to other different types of motions for texture videos, the distance information of the object in the scene from the camera surface, i.e. depth map, has to be used according to the image projection principle. Generally, depth map contains fewer details than texture, especially for the low-resolution case. Therefore, a Motion information based Coding scheme using Frame-Skipping for Depth map sequence (MCFSD) is proposed. The experimental results show that this scheme is effective for low resolution depth map sequences, which enhances the performance by around 2.0 dB. The idea of motion information assisted coding is finally employed to both texture sequence and depth map sequence for different types of motions. A Motion information based Texture plus Depth map Coding (MTDC) scheme is proposed for 3D videos. Moreover, this scheme is applied to H.264/AVC and the last H.265/HEVC video coding standard and tested for VGA resolution and HD resolution. The results show that the proposed scheme improves the performance under all the conditions. For VGA resolution under H.264/AVC standard, the average gain is about 2.0 dB. As the last H.265/HEVC enhances the video encoding efficiency, the average gain for HD resolution under H.265/HEVC standard drops to around 0.4 dB. Another contribution of this thesis is that a software plus hardware experimental data acquisition method is designed. The proposed motion information based video coding schemes require video sequences with accurate camera motion information. However, it is difficult to find proper dataset. Therefore, an embedded hardware based experimental data acquisition platform is designed to obtain real scene video sequences, while a CG based method is used to produce HD video sequences with accurate depth map.