Object Motion Research Articles

UAV Complex-Scene Single-Target Tracking Based on Improved Re-Detection Staple Algorithm

With the advancement of remote sensing technology, the demand for the accurate monitoring and tracking of various targets utilizing unmanned aerial vehicles (UAVs) is increasing. However, challenges such as object deformation, motion blur, and object occlusion during the tracking process could significantly affect tracking performance and ultimately lead to tracking drift. To address this issue, this paper introduces a high-precision target-tracking method with anomaly tracking status detection and recovery. An adaptive feature fusion strategy is proposed to improve the adaptability of the traditional sum of template and pixel-wise learners (Staple) algorithm to changes in target appearance and environmental conditions. Additionally, the Moth Flame Optimization (MFO) algorithm, known for its strong global search capability, is introduced as a re-detection algorithm in case of tracking failure. Furthermore, a trajectory-guided Gaussian initialization technique and an iteration speed update strategy are proposed based on sexual pheromone density to enhance the tracking performance of the introduced re-detection algorithm. Comparative experiments conducted on UAV123 and UAVDT datasets demonstrate the excellent stability and robustness of the proposed algorithm.

Deep progressive feature aggregation network for multi-frame high dynamic range imaging

High dynamic range (HDR) imaging is an important task in image processing that aims to generate well-exposed images in scenes with varying illumination. Although existing multi-exposure fusion methods have achieved impressive results, generating high-quality HDR images in dynamic scenes remains difficult. The primary challenges are ghosting artifacts caused by object motion between low dynamic range images and distorted content in underexposure and overexposed regions. In this paper, we propose a deep progressive feature aggregation network for improving HDR imaging quality in dynamic scenes. To address the issues of object motion, our method implicitly samples high-correspondence features and aggregates them in a coarse-to-fine manner for alignment. In addition, our method adopts a densely connected network structure based on the discrete wavelet transform, which aims to decompose the input features into multiple frequency subbands and adaptively restore corrupted contents. Experiments show that our proposed method can achieve state-of-the-art performance under different scenes, compared to other promising HDR imaging methods. Specifically, the HDR images generated by our method contain cleaner and more detailed content, with fewer distortions, leading to better visual quality.

Real-time uncertainty reduction in laser triangulation via dynamic speckle correlation

The speckle-induced measurement uncertainty is one of the major bottlenecks to achieve high measurement accuracy in laser triangulation sensors (LTS). Conventional methods for improving uncertainty focus on suppressing speckles and refining centroid extraction algorithms, overlooking the potential information about object motion that speckles inherently contain. To address this deficiency, a dynamic speckle correlation (DSC) method is proposed to reduce speckle-induced measurement uncertainty in laser triangulation systems. This method uses the correlation between speckle patterns and the movement of the object being measured, thereby enabling real-time error compensation without any speckle reduction mechanisms. To verify the effectiveness of the DSC method, we have designed a compact LTS based on oblique incident laser. Extensive experiments demonstrate that, for displacement measurement of objects with surface roughness ranging from Ra0.4 to Ra3.2, the DSC method can effectively reduce the root mean square (RMS) measurement error to 3∼5μm. Compared with using the traditional grayscale centroid method alone, the measurement accuracy after DSC compensation is improved by at least 20%.

Anomalous dynamics of a passive droplet in active turbulence

Motion of a passive deformable object in an active environment serves as a representative of both in-vivo systems such as intracellular particle motion in Acanthamoeba castellanii, or in-vitro systems such as suspension of beads inside dense swarms of Escherichia coli. Theoretical modeling of such systems is challenging due to the requirement of well resolved hydrodynamics which can explore the spatiotemporal correlations around the suspended passive object in the active fluid. We address this critical lack of understanding using coupled hydrodynamic equations for nematic liquid crystals with finite active stress to model the active bath, and a suspended nematic droplet with zero activity. The droplet undergoes deformation fluctuations and its movement shows periods of “runs” and “stays”. At relatively low interfacial tension, the droplet begins to break and mix with the outer active bath. We establish that the motion of the droplet is influenced by the interplay of spatial correlations of the flow and the size of the droplet. The mean square displacement shows a transition from ballistic to normal diffusion which depends on the droplet size. We discuss this transition in relation to spatiotemporal scales associated with velocity correlations of the active bath and the droplet.

Visualization of the rolling motion of spheres connected with a rubber band using an image processing technique

The traditional physics laboratory has adopted a photogate and polarized light to measure rotation motion in a fixed system, but they are not applicable to rolling objects. Computer vision has played an increasingly important role in image analysis. The popularization of mobile phones, cameras and computers makes it possible to measure the rotation of an object in a simple but nonintrusive way in a physics laboratory. A ‘tracker’ has been used to analyze the rotation motion of large objects, such as vehicle wheels, by recognizing the position of a marker. However, it is still challenging to precisely recognize the marker in a small object via simple algorithms in a physics laboratory. In this paper, we introduce a facile image processing technique to visualize the rotation and spin motion of spheres by marking it with two red and green semi-spheres. A basic formula of projection was derived to analyze the expression of spin angles with the projection area of the semi-sphere. The precision and suitability of the image processing technique was proven to be effective through comparison with the photogate measurement technique. The image processing technique has also been used to analyze the rotation and spin motion of two spheres bonded with a rubber band. It provides not only the trajectory of sphere centers, but the evolution of spin angles, orbit diameter and orbit angles of spheres. By analyzing the relationship between spin angles and orbit diameter, it is clear that the rubber band experienced typical transformations between three different phases (twist, coil and supercoil) until the energy was exhausted. The method offers valuable insights into rotational dynamics, showcasing its potential for practical applications in the physics laboratory and in educational contexts.

Observing gravitational waves with solar system astrometry

The subtle influence of gravitational waves on the apparent positioning of celestial bodies offers novel observational windows [1,2,3,4]. We calculate the expected astrometric signal induced by an isotropic Stochastic Gravitational Wave Background (SGWB) in the short distance limit. Our focus is on the resultant proper motion of Solar System objects, a signal on the same time scales addressed by Pulsar Timing Arrays (PTA). We derive the corresponding astrometric deflection patterns, finding that they manifest as distinctive dipole and quadrupole correlations or, in some cases, may not be present. Our analysis encompasses both Einsteinian and non-Einsteinian polarisations. We estimate the upper limits for the amplitude of SGWBs that could be obtained by tracking the proper motions of large numbers of solar system objects such as asteroids. We find that for SGWBs with negative spectral indices, such as that generated by Super Massive Black Hole Binaries (SMBHB), the constraints from these observations could rival those from PTAs. With the Gaia satellite and the Vera C. Rubin Observatory poised to track an extensive sample of asteroids — ranging from 𝒪(105) to 𝒪(106), we highlight the significant future potential for similar surveys to contribute to our understanding of the SGWB.

Identifikasi Konsep Fisika pada Bidang Miring Berbantuan Aplikasi Algodoo

This study aimed to determine the effectiveness of using the Algodoo application in the process of understanding the motion of objects descending an inclined plane. The method used in this research was an experimental method with a quantitative approach to identify the physical quantities of objects descending an inclined plane. The physical quantities were acceleration, velocity, force, potential energy, kinetic energy, and mechanical energy. Based on the data analysis, the acceleration of the object was found to be 4.9 m/s². The velocity of the object changed in a linear equation of 4.9t m/s. The potential energy decreased as the object descends the inclined plane. The kinetic energy of the object increased with the equation 1.2005t² J. The total of potential energy and kinetic energy, the mechanical energy, remained constant at 2.6956 J. These results indicated that Algodoo can be used to analyze physical quantities whose experimental values are close to theoretical ones. Therefore, Algodoo has the potential as an effective learning resource for understanding physics concepts.

Design and Application of Basketball Tactical Drill System Based on Motion Trajectory Analysis Algorithm

A basketball tactical drill system based on motion trajectory analysis algorithms employs advanced computational methods to enhance training effectiveness and strategy development. By utilizing these algorithms to analyze player movements and interactions during drills, coaches can gain valuable insights into team dynamics, spatial awareness, and decision-making processes. This analysis enables coaches to tailor drills to specific tactical objectives, such as improving defensive rotations, offensive spacing, or fast break execution. Additionally, the system can provide real-time feedback to players, helping them understand their positioning and movement patterns to refine their skills and decision-making abilities. This paper presents the design and implementation of a Basketball Tactical Drill System leveraging advanced object detection and motion trajectory analysis algorithms. The system aims to enhance basketball training by providing coaches and players with real-time insights into tactical drills and player performance. Two key algorithms, Mean-Shift and Difference Object Detection with Motion Trajectory (DOD-MT), are utilized to accurately identify and track basketballs, players, and other relevant objects during training sessions. Experimental results demonstrate the effectiveness of the proposed system, showcasing high detection accuracy, minimal false positives, and competitive processing times across various scenarios and tactical drills. The system leverages Mean-Shift and Difference Object Detection with Motion Trajectory (DOD-MT) algorithms to achieve exceptional detection accuracy, with an average of 93% across various scenarios and tactical drills. False positives are minimized, with an average of 12 per scenario. Processing times remain competitive, averaging 27 milliseconds per scenario. Integration of motion trajectory analysis provides coaches with invaluable real-time insights into player movements and interactions, facilitating data-driven coaching strategies. The system's versatility is demonstrated through successful application in a range of drills, resulting in performance improvements averaging 20%.

OBT-Trace: An approach to trace and recognition object motion through prosthetic hand gestures

The development of prosthetic hands has advanced significantly in recent years, aiming to provide more intuitive and functional solutions for individuals with upper limb amputations. This research presents a novel approach to prosthetic hand control by integrating 3D hand gesture recognition with object manipulation and recognition capabilities. Our proposed system utilizes a pre-trained object recognition model, based on transfer learning, to enable the prosthetic hand to perceive and identify objects in its vicinity. The model leverages a vast dataset of objects, enabling the prosthetic hand to recognize a wide array of everyday items, thus enhancing its versatility.In addition, the prosthetic hand incorporates a sophisticated 3D hand gesture recognition system, allowing users to control the hand's movements and actions seamlessly. By recognizing specific gestures, such as grasping, lifting, and releasing, users can intuitively interact with their environment and perform various tasks with ease. This research leverages the synergy between gesture recognition and object recognition, creating a powerful framework for prosthetic hand control. The system's adaptability and versatility make it suitable for a broad range of applications, from assisting with daily tasks to enhancing the quality of life for individuals with upper limb amputations.The results of this study demonstrate the feasibility and effectiveness of combining 3D hand gesture recognition with pre-trained object recognition through transfer learning. This approach opens up new possibilities for enhancing prosthetic hand functionality and usability, ultimately improving the lives of those who rely on these devices for daily living. The proposed model combines the features of YOLO V7 object detection with pre-trained models. The proposed model achieves 99.8% of accuracy compared to the existing models.

The hydrodynamic study of the interaction between waves and objects in permeable reef environment

This study employed a self-developed viscous numerical wave tank based on the piston type wave maker and user-defined functions of Fluent software to explore the wave shadowing effect of the transmission of waves near the underwater object and the reef. In the study, various factors such as the permeability of the reef, changes in topography, and objects motion were considered, and detailed simulation analyses were conducted for five scenarios: a solitary reef, a fixed underwater object, a forced moving underwater object, and a free moving underwater object. The results showed that this method can effectively simulate the complex flow field changes around the underwater object and the reef under different conditions, while also fully considering the strong nonlinearity and fluid viscosity effects between the wave surface and the object. Furthermore, it was found that between the free-moving object and the reef, there would be significant swirls and streamline distortion. This indicates that a portion of the incident wave energy is transformed into kinetic energy of swirls, resulting in a smoother and more uniform flow field and reducing wave forces.

Meter‐Scale Distance Manipulation of Diverse Objects with Jet‐Induced Airflow Field

In nature, wind can transport objects of diverse shapes and materials over long distances spanning meters or beyond. In contrast, most noncontact manipulation methods are effective only over centimeter‐scale distances and require the manipulated objects to have specific magnetic, electrical, or other material properties and shapes. Herein, a meter‐scale distance manipulation method is presented for controlling the motion of objects of diverse shapes and materials on a plane surface using a jet‐induced airflow field. The airflow field is varied by controlling the direction of a single air jet projected onto a surface based on the positions of objects to steer the objects to follow desired trajectories up to 2.7 m away. A wide variety of objects can be automatically manipulated, from regularly shaped polystyrene hemispheres and sticks to irregularly shaped cotton wads and face masks as well as deformable tissue papers and plastic bags. The method is also robust, effectively manipulating objects under challenging conditions, e.g., external airflow disturbances, surface obstacles, and air–water interface. Finally, three application cases are demonstrated: collecting diverse objects into a target receptacle, hooking and retrieving heavy objects with a tethered mobile agent, and closing an electrical circuit with an untethered soft agent.

CROWD VIDEO SEQUENCES PROCESSING METHODS FOR DETERMINING THE CROWD MOTION PATTERNS

Nowadays, homogeneous objects clusters motion is one of the most important and rapidly developing computer vision and machine learning application. In this paper, we consider the crowd motion patterns determination by using motion maps that we calculate with FlowNet, a neural network examining motion of objects in a video sequence. This approach allows us to get information on the crowd direction and velocity with relation to other objects of scene, which plays the key role in behavior analysis and security establishment. Besides, we consider methods for preliminary video sequence processing, including frame combination, to estimate motion maps more precisely and improve the effectiveness of the dynamic scenes analysis.

Analytic Solutions for the Motion of Spinning Particles near Spherically Symmetric Black Holes and Exotic Compact Objects.

Rapidly rotating bodies moving in curved space-time experience the so-called spin-curvature force, which becomes important for the motion of compact objects in gravitational-wave inspirals. As a first approximation, this effect is captured in the motion of a spinning test particle. We solve the equations motion of a spinning particle to leading order in spin in arbitrary static and spherically symmetric space-times in terms of one-dimensional closed-form integrals. This solves the problem and proves its integrability in a wide range of modified gravities and near exotic compact objects. Then, by specializing to the case of bound orbits in Schwarzschild space-time, we demonstrate how to express the solution in the form of Jacobi elliptic functions.

Numerical Simulation and Analysis of Added Mass for the Underwater Variable Speed Motion of Small Objects

Unlike uniform motion, when an object moves underwater with variable speed, it experiences additional resistance from the water, commonly referred to as added mass force. At present, several methods exist to solve this force, including theoretical, experimental, and simulation approaches. This paper addresses the challenge of determining the added mass force for irregularly shaped small objects undergoing variable speed motion underwater, proposing a method to obtain the added mass force through numerical simulation. It employs regression analysis and parameter separation analysis to solve the added mass force, added mass, viscous drag coefficient, and pressure drag coefficient. The results indicate that an added mass force exists during both the acceleration and deceleration of the object, with little difference between them. Under the same velocity conditions, significant differences exist in pressure drag forces, while differences in viscous drag forces are not significant. This suggests that the primary source of added mass force is pressure drag, with viscous drag having little effect on it. During acceleration, the surrounding fluid accelerates with the object, increasing the pressure drag with a high-pressure area concentrating at the object’s front, forming an added mass force that is directed backward. By contrast, during deceleration, the fluid at the object’s front tends to detach, and the fluid at the rear rushes forward, leading to a smaller high-pressure area at the front and a larger one at the rear, reducing the pressure drag and forming an added mass force that is directed forward. By comparing the added mass of a standard ellipsoid obtained from numerical simulation with theoretical values, the regression analysis method is proven to be highly accurate and entirely applicable for solving the added mass of underwater vehicles.

A novel artificial visual system for motion direction detection in color images

In recent years, convolutional neural networks (CNNs) have dominated the field of computer vision. Compared to traditional methods, these neural network algorithms exhibit strong biomimetic performance advantages in complex visual tasks due to their brain-like structure. However, because some necessary neural characteristics are ignored, these neural network algorithms differ greatly from the computational mechanisms of the brain. This paper starts with extracting basic visual features such as motion direction information from the brain and abstracts, generalizes and models a novel artificial visual system (AVS) for detecting object motion direction in color images based on existing relevant neurophysiological knowledge. We propose a mathematical model and quantification mechanism for each component neuron that generates motion direction selectivity in the visual system using dendritic neuron models, spiking neural network concepts and neurophysiological knowledge of retinal direction-selective ganglion cell pathways. The experiment is based on one million instances of object motion under different environments of noise-free, static and dynamic random noise, dynamic salt-and-pepper noise, dynamic Gaussion noise, and dynamic light changing. In comparison with 4 famous CNNs, LeNet-5, EfficientNetB0, ResNet18, and RegNetX-200MF, we test and verify AVS’s effectiveness, efficiency and strong generalization ability as well as other biomimetic performance advantages including high biological rationality, learning-free capability, interpretability and ease-of-use etc. AVS demonstrates that neuroscience still has important implications for guiding and promoting the development of artificial intelligence technology. Furthermore, AVS firstly provides a successful quantitative reference case study for further understanding motion direction selectivity and other primary cortical encoding characteristics in the brain.

Atypical Time to Contact Estimation in Young Adults with Autism Spectrum Disorder.

Individuals with Autism Spectrum Disorder (ASD) present atypical sensory processing in the perception of moving stimuli and biological motion. The present study aims to explore the performance of young adults with ASD in a time to contact (TTC) estimation task involving social and non-social stimuli. TTC estimation involves extrapolating the trajectory of a moving target concealed by an occluder, based on the visible portion of its path, to predict the target's arrival time at a specific position. Sixteen participants with a diagnosis of level-1 ASD (M = 19.2 years, SE = 0.54 years; 3F, 13M) and sixteen participants with TD (M = 22.3 years, SE = 0.44 years; 3F, 13M) took part in the study and underwent a TTC estimation task. The task presented two object types (a car and a point-light walker), different object speeds, occluder lengths, motion directions and motion congruency. For the car object, a larger overestimation of TTC emerged for ASDs than for TDs, whereas no difference between ASDs and TDs emerged for the point-light walker. ASDs exhibited a larger TTC overestimation for the car object than for the point-light walker, whereas no difference between object types emerged for TDs. Our results indicated an atypical TTC estimation process in young adults with ASD. Given its importance in daily life, future studies should further explore this skill. Significant effects that emerged from the analysis are discussed.

Three-dimensional wireless measurement of float–sink object motions in a gas–solid fluidized bed

We developed a Lagrangian sensor system capable of wirelessly measuring the 3D motions of objects and applied it to examine the floating and sinking behaviors in a gas–solid fluidized bed. The system consisted of a sensor particle, comprising a 9-axis sensor and a wireless module integrated into a spherical outer shell, and three pairs of magnetic coils. By utilizing this system, we can non-invasively measure the 3D motions of objects, which cannot be observed from the outside. In this study, we examined the motions of objects with various densities in the gas–solid fluidized bed. We found the similarity and dissimilarity between the float–sink motions of objects in fluidized beds and those in liquids. The experimental results revealed that the vertical motion of an object in the fluidized bed depends on its density and can be categorized into two distinct states: floating near the surface or settling down at the bottom. However, within a specific range of object densities close to the apparent density of the fluidized bed, the object exhibited complicated motions in the bed. In such cases, slight density variations induced unpredictable changes between the floating and sinking states. The results also suggested that the settling velocity of the objects in the fluidized bed varies with their vertical positions in the bed.

Utilization of Video-based Laboratory (VBL) Using Tracker for Analysis of Object Motion on the Laboratory-Scaled Mini Roller Coaster

The use of teaching aids in learning can provide direct learning experiences to students so that learning becomes more meaningful. Students observe physics phenomena that occur and get physics data from these phenomena. The observed physics phenomena often happen very quickly, so it is impossible to record data manually. Therefore, a method is needed to facilitate data recording and analysis. One way to use a video-based laboratory (VBL) is by utilizing a tracker. This method allows for analyzing the objects contained in a video and displays the data in graphs and tables. This study aims to investigate the motion of objects using tracker video analysis on a mini-roller coaster tool. The video analyzed is a video of a solid ball-shaped object with a mass of 8.6 g moving along a roller-coaster trajectory. By tracking an object, physics data will be obtained to analyze acceleration and the law of conservation of mechanical energy. The analysis results match the experimental results and the physics concepts that apply to the roller coaster motion.
 Keywords: video-based laboratory (VBL), tracker, laboratory-scaled mini roller coaster

Estimate and compensate head motion in non-contrast head CT scans using partial angle reconstruction and deep learning.

Patient head motion is a common source of image artifacts in computed tomography (CT) of the head, leading to degraded image quality and potentially incorrect diagnoses. The partial angle reconstruction (PAR) means dividing the CT projection into several consecutive angular segments and reconstructing each segment individually. Although motion estimation and compensation using PAR has been developed and investigated in cardiac CT scans, its potential for reducing motion artifacts in head CT scans remains unexplored. To develop a deep learning (DL) model capable of directly estimating head motion from PAR images of head CT scans and to integrate the estimated motion into an iterative reconstruction process to compensate for the motion. Head motion is considered as a rigid transformation described by six time-variant variables, including the three variables for translation and three variables for rotation. Each motion variable is modeled using a B-spline defined by five control points (CP) along time. We split the full projections from 360° into 25 consecutive PARs and subsequently input them into a convolutional neural network (CNN) that outputs the estimated CPs for each motion variable. The estimated CPs are used to calculate the object motion in each projection, which are incorporated into the forward and backprojection of an iterative reconstruction algorithm to reconstruct the motion-compensated image. The performance of our DL model is evaluated through both simulation and phantom studies. The DL model achieved high accuracy in estimating head motion, as demonstrated in both the simulation study (mean absolute error (MAE) ranging from 0.28 to 0.45mm or degree across different motion variables) and the phantom study (MAE ranging from 0.40 to 0.48mm or degree). The resulting motion-corrected image, , exhibited a significant reduction in motion artifacts when compared to the traditional filtered back-projection reconstructions, which is evidenced both in the simulation study (image MAE drops from 178 33HU to 37 9HU, structural similarity index (SSIM) increases from 0.60 0.06 to 0.98 0.01) and the phantom study (image MAE drops from 117 17HU to 42 19HU, SSIM increases from 0.83 0.04 to 0.98 0.02). We demonstrate that using PAR and our proposed deep learning model enables accurate estimation of patient head motion and effectively reduces motion artifacts in the resulting head CT images.

Self‐Powered Multimodal Sensing Using Energy‐Generating Solar Skin for Robotics and Smart Wearables

Wearable electronic devices‐laden systems such as electronic‐skin (e‐Skin) have been explored in recent years to enable advances in applications such as Internet of Things, healthcare, and robotics. The power requirement of multitudes of devices in the e‐Skin is a major hurdle for its wider uptake. Herein, a solar cells‐based energy generating e‐Skin is presented and how the energy outputs of solar cells can be innovatively processed for multimodal sensing is demonstrated. By reading the variations and energy output patterns of the e‐Skin, present on a robotic arm, multiple parameters can be sensed including object motion, color detection, and ambient temperature. With the accurate tracking of shadow sensing, for an object moving in horizontal and vertical directions with respect to the solar skin, information can be obtained such as the velocity and acceleration of moving object. In this regard, the presented e‐Skin can also be seen to have vision capability. The presented multifunctional energy‐generating e‐Skin shows an energy surplus of >1 mW (effective module area of 20 cm2) under white light illumination of 4,450 lux, which is sufficient for continuous powering of portable low‐powered devices. Finally, we demonstrate the e‐Skin application for energy‐autonomous hand gestures recognition in robotics.