Object Motion Research Articles

FLODCAST: Flow and depth forecasting via multimodal recurrent architectures

Forecasting motion and spatial positions of objects is of fundamental importance, especially in safety-critical settings such as autonomous driving. In this work, we address the issue by forecasting two different modalities that carry complementary information, namely optical flow and depth. To this end we propose FLODCAST a flow and depth forecasting model that leverages a multitask recurrent architecture, trained to jointly forecast both modalities at once. We stress the importance of training using flows and depth maps together, demonstrating that both tasks improve when the model is informed of the other modality. We train the proposed model to also perform predictions for several timesteps in the future. This provides better supervision and leads to more precise predictions, retaining the capability of the model to yield outputs autoregressively for any future time horizon. We test our model on the challenging Cityscapes dataset, obtaining state of the art results for both flow and depth forecasting. Thanks to the high quality of the generated flows, we also report benefits on the downstream task of segmentation forecasting, injecting our predictions in a flow-based mask-warping framework.

Human-robot planar co-manipulation of extended objects: data-driven models and control from human-human dyads.

Human teams are able to easily perform collaborative manipulation tasks. However, simultaneously manipulating a large extended object for a robot and human is a difficult task due to the inherent ambiguity in the desired motion. Our approach in this paper is to leverage data from human-human dyad experiments to determine motion intent for a physical human-robot co-manipulation task. We do this by showing that the human-human dyad data exhibits distinct torque triggers for a lateral movement. As an alternative intent estimation method, we also develop a deep neural network based on motion data from human-human trials to predict future trajectories based on past object motion. We then show how force and motion data can be used to determine robot control in a human-robot dyad. Finally, we compare human-human dyad performance to the performance of two controllers that we developed for human-robot co-manipulation. We evaluate these controllers in three-degree-of-freedom planar motion where determining if the task involves rotation or translation is ambiguous.

Identification of the object motion trajectory in a digital image using the Kalman filter

The paper concerns identification of the object motion trajectories in a digital image. An image analysis procedure was used consisting of two methods: BLOB identification and Kalman filter. BLOB identification is the superior method, and the Kalman filter is an alternative one, used to determine the motion trajectories only in the event of incorrect BLOB identification. Data from the correct BLOB identification are used on an ongoing basis in the process of training the Kalman filter. In this way, the filter becomes a model of a physical object that allows predicting the future state based on the object's current behaviour. The use of two complementary image analysis methods allows to determine the motion trajectories of the object's characteristic points in conditions of uneven scene lighting. The effectiveness of the method was confirmed in experimental rests.

Scalar fields features of video sequences energy characteristics

To improve the methods of spatio-temporal filtering, an approach based on the analysis of finitary images energy spectra can be used. The purpose of this work is to study the possibility of using energy spectra for motion identification in video sequence frames. Each frame of a video sequence is a spatially finite image, the energy spectrum of which contains information about the phase components. This information is of particular importance for determining the location of dynamic objects. The most important information is also contained in the harmonic amplitude increments of the frame energy spectra, i.e. the energy spectrum generates a scalar discrete field. The energy functions describe inter-frame changes in the amplitudes of harmonics of the energy spectrum. To apply energy spectra in motion identification algorithms, it is necessary to analyze the properties of these scalar fields. Scalar fields generated by energy spectra of frames of test video sequences are studied. Test video sequences with an object moving in an arbitrary direction are used. Scalar fields of energy functions are presented graphically. Characteristic regions of scalar fields are identified and their projections are built using isolines. The boundaries of the characteristic areas and the coordinates of extrema of energy functions are determined. From them, all coordinates of the dynamic object and its horizontal/vertical displacements can be found. The dependence of phase-energy function extrema on the position of the object before the motion and on the direction of motion is studied. When moving in an arbitrary direction, different types of extrema are observed: point and extended. At vertical or horizontal movements, the energy functions graphs have only two-dimensional pulses. Additional criteria for eliminating ambiguities are proposed. The ambiguities are due to the fact that the number and nature of extrema may coincide for different directions of motion and location of objects in different quadrants of the coordinate plane. The extrema of energy functions for object movements in all possible directions and locations in any quadrant of a video sequence frame are determined. The obtained results showed that energy functions for object motion in arbitrary direction have a large number of characteristic regions containing point and extended extrema. The configuration of extrema depends both on the direction of motion and on the quadrant of the image in which the object is located. Therefore, it is most advisable to begin to determine the spatial coordinates of an object and the magnitude of its horizontal/vertical movement by finding the maximum and minimum boundaries of energy functions, and then refine it using the coordinates of the extrema. Consideration of characteristic regions and extrema of scalar fields of energy functions can be used in the development of algorithms effective in terms of "accuracy/speed" ratio, and can increase the performance of image and video processing systems. Algorithms for determining the object position, velocity and direction of its movement by analyzing scalar fields of energy spectra components can be used as an additional or main tool for motion identification in technical vision systems.

Kinematics analysis on accelerated motion using tracker video analysis for educational purposes

This study aims to determine the effectiveness of video analysis tracker-based practicum on the material of the motion of objects on an inclined plane and the motion of objects in free fall. The use of video tracker analysis-based practicum is useful for determining some kinematics parameters of objects that cannot only be calculated by traditional practicum methods. By using the video tracker analysis method, all object kinematics parameters can be found directly in visual form. The results show that (1) There is no relationship between mass and kinematics parameters. (2) There is an influence due to air friction on the acceleration value of the object. (3) The suitability of the concept with the tracker application produced with the error value in the experiment of object motion on an inclined plane () (4) The suitability of the concept with the tracker application produced with the error value in the experiment of object motion in free fall () (5) The graphical representation of the tracker is in accordance with the theory studied. This research implies that the existence of learning using practicum, especially practicum-based video tracker analysis can improve understanding and process skills in understanding science concepts. This research is written so that the programme can be further applied to science education.

Moving object detection algorithm and motion capture based on 3D LiDAR

Moving object detection algorithm and motion capture based on 3D LiDAR

Aristotle’s Static World and Traditional Education – A Secondary Publication

The purpose of this study is to understand the characteristics of Aristotle’s view of nature that is, the staticview of the universe, and find implications for education. Plato sought to interpret the natural world using a rationalapproach rather than an incomplete observation. He thought that viewing the world from the perspective of geometry andmathematical regularity was the best way to understand the world. On the other hand, Aristotle believed that we couldunderstand the world through observations. His worldview portrayed a static reality, imbued with individual purpose anda sense of deliberate legitimacy. In addition, the natural motion of earthly objects and celestial bodies, which are naturalmovements towards the world of order, are the original actions. Aristotle thought that all natural things would carry outsome movement. Above all, Plato and Aristotle view the world as a static universe. It is possible to fully grasp the worldby approaching the objective nature that exists independently of human beings with human reason and observation.After all, for Aristotle, like Plato, their belief that the natural world was subject to regular and orderly laws of nature,despite the complexity and seemingly unceasing changes, became the basis of Western thought. The universe, rooted inthe metaphysical perspective of ancient Greece and carried forward in modern philosophy, has traditionally depended ondeveloping a dichotomy of comprehension, distinguishing between what is already accomplished or predetermined, theideal and the inevitable. This dichotomy forms the foundation of traditional teaching methods that often do not prioritizethe opinions of learners.

Fully Memristive Elementary Motion Detectors for a Maneuver Prediction.

Insects can efficiently perform object motion detection via a specialized neural circuit, called an elementary motion detector (EMD). In contrast, conventional machine vision systems require significant computational resources for dynamic motion processing. Here, a fully memristive EMD (M-EMD) is presented that implements the Hassenstein-Reichardt (HR) correlator, a biological model of the EMD. The M-EMD consists of a simple Wye (Y) configuration, including a static resistor, a dynamic memristor, and a Mott memristor. The resistor and dynamic memristor introduce different signal delays, enabling spatio-temporal signal integration in the subsequent Mott memristor, resulting in a direction-selective response. In addition, a neuromorphic system is developed employing the M-EMDs to predict a lane-changing maneuver by vehicles on the road. The system achieved a high accuracy (> 87%) in predicting future lane-changing maneuvers on the Next Generation Simulation (NGSIM) dataset while reducing the computational cost by 92.9% compared to the conventional neuromorphic system without the M-EMD, suggesting its strong potential for edge-level computing.

Research on impact resistance of new suspended elastic safety net system

In order to reduce the direct injuries caused by falling from height, we propose a new type of suspended elastic safety net system suitable for pylon assembly construction. Analyze the correlation between damping and toughness of the protection system, establish the falling object motion control method, and construct the mechanism model of the protection system accordingly. Full-scale impact tests are carried out using test pylons on transmission line projects. Test data is collected using high-speed cameras and dynamic data acquisition equipment to analyze the key impact dynamic response characteristics of the test model. The experimental results show that the system has three working stages. In the first stage, the elastic safety net is transformed from an initial equilibrium state, where it is subjected only to its gravity, to a taut state. In the second stage, as the elongation of the system continues to increase in the direction of impact, the internal forces transmitted by the net to the elastic rod and the supporting rope gradually increase until the energy dissipating devices connected to the elastic rod are activated. In the third stage, additional elongation of the system is provided by activating the support springs in the bearing supports. Based on the failure modes observed in the tests, the key factors leading to the breakdown of the system are investigated. A new numerical simulation method is developed which is not only capable of predicting the load-bearing capacity of an elastic safety net system under the impact of a falling object, but also accurately simulates the main operating characteristics, such as buckling of the elastic rods, activation of the energy-absorbing devices, and deformation of the net ropes. The numerical simulation model was evaluated using the experimental results, and the numerical response of the system was in good agreement with the experimental data.

Dipolar skyrmions and antiskyrmions of arbitrary topological charge at room temperature.

Magnetic skyrmions are localized, stable topological magnetic textures that can move and interact with each other like ordinary particles when an external stimulus is applied. The efficient control of the motion of spin textures using spin-polarized currents opened an opportunity for skyrmionic devices such as racetrack memory and neuromorphic or reservoir computing. The coexistence of skyrmions with high topological charge in the same system promises further possibilities for efficient technological applications. In this work, we directly observe dipolar skyrmions and antiskyrmions with arbitrary topological charge in Co/Ni multilayers at room temperature. We explore the dipolar-stabilized spin objects with topological charges of up to 10 and characterize their nucleation process, their energy dependence on the topological charge and the effect of the material parameters on their stability. Furthermore, our micromagnetic simulations demonstrate spin-transfer-induced motion of these spin objects, which is important for their potential device application.

Development of Ethno-STEM-Loaded Digital Science Teaching Materials the Process of Making Traditional Sidoarjo Snacks Material of Force and Object Motion to Train Science Literacy in Elementary School Students

This study aims to develop digital science teaching materials with Ethno-STEM content to train elementary students' scientific literacy. The research method used is the ADDIE Research and Development (R&D) model (Analyze, Design, Development, Implementation and Evaluation). The results of the feasibility of digital science teaching materials obtained a percentage of 92.42% with very feasible criteria. The effectiveness of digital science teaching materials was declared effective by the results of the paired t-test which showed that there was an average difference in the pretest and posttest data, and the n-gain results scored 0.40 with the criteria of being effective. The results of the student response questionnaire after the learning process obtained a percentage of 90.35%; as well as the results of the teacher's response questionnaire after the learning process obtained a percentage of 90.62%. The results of student and teacher responses obtained very positive criteria. The conclusions of this study are that digital science teaching materials with Ethno-STEM content are effective in training scientific literacy for grade IV elementary school students.

Far-field speckle correlations as a function of object position for microscopically distinguishing objects hidden in a randomly scattering medium

Super-resolution optical sensing is of critical importance in science and technology and has required prior information about an imaging system or obtrusive near-field probing. Additionally, coherent imaging and sensing in heavily scattering media such as biological tissue has been challenging, and practical approaches have either been restricted to measuring the field transmission of a single point source, or to where the medium is thin. We present the concept of far-subwavelength spatial sensing with relative object motion in speckle as a means to coherently sense through heavy scatter. Experimental results demonstrate the ability to distinguish nominally identical objects with nanometer-scale translation while hidden in randomly scattering media, without the need for precise or known location and with imprecise replacement. The theory and supportive illustrations presented provide the basis for super-resolution sensing and the possibility of virtually unlimited spatial resolution, including through thick, heavily scattering media with relative motion of an object in a structured field. This work provides enabling opportunities for material inspection, security, and biological sensing.

Random vector representation of continuous functions and its applications in quantum mechanics

The relation between continuous functions and random vectors is revealed in the paper that the main meaning is described as, for any given continuous function, there must be a sequence of probability spaces and a sequence of random vectors where every random vector is defined on one of these probability spaces, such that the sequence of conditional mathematical expectations formed by the random vectors uniformly converges to the continuous function. This is a random vector representation of continuous functions, which is regarded as a bridge to be set up between real function theory and probability theory. By means of this conclusion, an interesting result about function approximation theory can be obtained. The random vector representation of continuous functions is of important applications in physics. Based on the conclusion, if a large proportion of certainty phenomena can be described by continuous functions and random phenomena can also be described by random variables or vectors, then any certainty phenomenon must be the limit state of a sequence of random phenomena. Then, in the approximation from a sequence of random vectors to a continuous function, the base functions are appropriately selected by us, and an important conclusion for quantum mechanics is deduced: classical mechanics and quantum mechanics are unified. Particularly, an interesting and very important conclusion is introduced as the fact that the mass point motion of a macroscopical object possesses a kind of wave characteristic curve, which is called wave-mass-point duality.

High-accuracy image-free classification of high-speed rotating objects with fluctuating rotation periods

Image-free classification methods with single-pixel measuring and deep learning show a capacity for long-duration classification of moving objects. However, motion blur restricts the allowable object motion speed of existing image-free classification methods. Aimed at high-speed rotating objects, we propose an image-free classification approach based on single-pixel measuring at the same spatial position of different rotation cycles to reduce motion blur. The proposed approach allows classifying rotating objects with fluctuating rotation periods to better meet the actual application conditions. We verify the proposed method by classifying the digits placed on a rotating disk. In our experiments, when digits rotate at around 960.9 revolutions per minute, corresponding to 10.06 m/s, the classification accuracy reaches 95.9%. In theory, the allowable speed is determined only by the sampling rate of the single-pixel measurements, which can allow for higher speeds than experimentally achieved. The proposed image-free classification method provides a promising way of monitoring high-speed rotating objects in the engineering field.

Three dimensional tracking of rigid objects in motion using 2D optical flows

Detecting and tracking objects in image sequences is paramount for any video analytic. While object detectors have become increasingly robust, motion estimates based on the straightforward approach of running a detector and linking detections are prone to bounding box noise. An algorithm for the monocular estimation of the 3D motion of rigid objects is presented, combining an object detector, minimal camera calibration, and 2D optical flows observed on the image plane. The algorithm utilizes the 2D Delaunay triangulation over geometrically consistent optical flow tracks to identify regions common to the same object. The algorithm is evaluated on the special case of image sequences of vehicles on roads. Experiments on BrnoCompSpeed dataset show that speed estimates from both detector-based tracking (mean error of 1.62 km/h) and optical flow based tracking (mean error of 2.19 km/h) perform competitively on straight roads. An extensive empirical evaluation is also conducted on a new dataset containing synthetic and real world scenes that also include vehicle trajectories involving rotation. A naive baseline bounding box track based method obtained a mean error of 5.29 km/h for vehicle speed, but optical flow tracks performed significantly better with mean error of 1.65 km/h on the new dataset including rotation. With the new method, video analytics such as vehicle speed estimation and lane change detection can obtain precise information about the 3D trajectory of rigid objects in motion.

Building the Largest Sample of IR Galaxies. Preliminary Results

To build the IRAS full extragalactic sample (including both point-like sources - QSOs, and extended ones - galaxies), we have applied several approaches. Cross-correlation of IRAS PSC/FSC Combined Catalogue with optical catalogues of already known galaxies, quasars and blazars such as The NGC 2000.0 Catalog, Third Reference Catalogue of Bright Galaxies, The Roma BZCAT, Astrometric Catalogue 5 and many others. Cross-correlation of IRAS PSC/FSC Combined Catalogue with optical catalogues giving data, which can be used to determine galaxy candidates, for example Sloan Digital Sky Survey (SDSS, contains data about point like and extended objects), The APM North Catalogue (contains data about ellipticity of objects), etc. Cross-correlation of IRAS PSC/FSC Combined Catalogue with GAIA DR3, which gives data about proper motions of objects (we can consider the objects having real proper motions as stars and exclude them from the sample). If all data show the same type of object, then we give it as a genuine one, and if there is an ambiguity, we give the most probable type with a flag. All this will lead to construction of a confident sample of optical counterparts of extragalactic objects of IRAS PSC/FSC: galaxies and quasars. We also will study IR/opt flux ratios, which may serve as one more characteristic to reveal galaxies with very high SFR. We will carry out calculation of IR luminosities for all IR galaxies. We expect revelation of many new ULIRGs and HLIRGs and building their largest samples.

GLOCAL: A self-supervised learning framework for global and local motion estimation

Motions in videos are typically a mixture of local dynamic object motions and global camera motion, which are inconsistent in some cases, and even interfere with each other, causing difficulties in various downstream applications, such as video stabilization that requires the global motion, and action recognition that consumes local motions. Therefore, it is crucial to estimate them separately. Existing methods separate two motions from the mixed motion fields, such as optical flow. However, the quality of mixed motion determines the higher bounds of the performance. In this work, we propose a framework, GLOCAL, to directly estimate global and local motions simultaneously from adjacent frames in a self-supervised manner. Our GLOCAL consists of a Global Motion Estimation (GME) module and a Local Motion Estimation (LME) module. The GME module involves a mixed motion estimation backbone, an implicit bottleneck structure for feature dimension reduction, and an explicit bottleneck for global motion recovery based on the global motion bases with foreground mask under the training guidance of proposed global reconstruction loss. An attention U-Net is adopted for LME which produces local motions while excluding motion of irrelevant regions under the guidance of proposed local reconstruction loss. Our method can achieve better performance than the existing methods on the homography estimation dataset DHE and the action recognition dataset NCAA and UCF-101.

Black holes shielded by magnetic fields

Black holes (BHs) formed by collapsing and/or merging of magnetized progenitors, have magnetic fields penetrating the event horizon, and there are several possible scenarios. Thus, the no-hair theorem that assumes the outside medium is a vacuum, is not applicable in this case. Bearing this in mind and considering a Schwarzschild BH of mass M immersed in a uniform magnetic field B, we show that all three frequencies related to the equatorial circular orbit of a test particle become imaginary for the orbits of radii rB>2B−1. It signifies that if a BH is surrounded by a magnetic field of order B∼Rg−1 (where Rg is the gravitational radius of the BH), a test particle could unable to continue its regular geodesic motion from/at r>rB, hence the accretion disk could not be formed, and the motion of other stellar objects around the BH could be absent. As the BHs are generally detected by watching for their effects on nearby stars and gas, a magnetic field of order B∼Rg−1 could be able to shield a BH in such a way that it could remain undetectable. Motivated with this theoretical investigation and considering the sphere (of radius rf) of magnetic influence around an astrophysical BH, we constrain B, above which a magnetized BH could remain undetectable. For example, M=109M⊙ BH surrounded by B>106 G and M=10M⊙ BH surrounded by B>1014 G could remain undetectable for rf∼105Rg. In other words, our result also explains why a detected SMBH has surprisingly weak magnetic field.

Developing an Interactive Simulation for Noninertial Reference Frames

The forces involved in the motion of objects within noninertial reference frames are challenging concepts for introductory and advanced mechanics students. Furthermore, students often struggle to predict the trajectories of an object due to these fictitious forces. While most students encounter noninertial reference frames daily, students often spend most of their physics coursework analyzing situations in which the Coriolis and centrifugal forces are negligible. We developed an interactive and freely accessible simulation of a rotating turntable apparatus that aims to improve understanding of the motion of objects in noninertial reference frames.

DLOAM: Real-time and Robust LiDAR SLAM System Based on CNN in Dynamic Urban Environments

Dynamic object detection, state estimation, and map-building are crucial for autonomous robot systems and intelligent transportation applications in urban scenarios. Most current LiDAR Simultaneous Localization and Mapping (SLAM) systems operate on the assumption that the observed environment is static. However, the overall accuracy and robustness of a SLAM system can be compromised by dynamic objects in the environment. Aiming at the problem of inaccurate odometry estimation and wrong mapping caused by the existing LiDAR SLAM method which cannot detect the dynamic objects, we study the SLAM problem of robots and unmanned vehicles equipped with LiDAR traveling in the dynamic urban scenes. We propose a fast LiDAR-only model-free dynamic objects detection method, which uses the spatial and temporal information of point cloud through a convolutional neural network (CNN), and the detection accuracy is improved by 35 use spatial information. We further integrate it into a state-of-the-art LiDAR SLAM framework to improve the SLAM performance. Firstly, the range image constructed by LiDAR point cloud is used for ground extraction and non-ground point clustering. Then, the motion of objects in the scene is estimated by the difference between adjacent frames, and the segmented objects are further divided into dynamic objects and static objects by their motion features. After that, the stable feature points are extracted from the static objects. Finally, the pose transformation of adjacent frames is solved by matching feature point pairs. We evaluated the accuracy and robustness of our system on datasets with different challenging dynamic environments, and the results show our system has significant improvements in accuracy and robustness of odometry and mapping, while still maintain real-time performance, which is sufficient for autonomous robot systems and intelligent transportation applications in urban scenarios.