Search

Due to the high cost and large error of traditional UAV big data tracking and automatic measurement technology, a method of big data tracking and automatic measurement for UAV trajectory based on MEMS sensor was put forward. The iterative learning control algorithm was used to estimate the repetitive disturbance and modeling error of system based on the simplified dynamics model of four-rotor helicopter and the optimal estimation characteristics of Kalman filter. The discrete equation of quadratic performance function in time domain was selected to compensate the estimated model error disturbance, and then the big data tracking was completed. Based on the data of gyroscope, the quaternion differential equation was established. The differential equation was solved by first-order Picard method, and a set of quaternion data was obtained. The gradient descent method was used to process the acceleration data and magnetic data, and thus to get the optimal quaternion. Finally, the measurement results were obtained by fusing the two quaternions with MEMS sensors. Simulation results prove that the proposed method can obtain the trajectory tracking data and measurement information of UAV accurately.

This paper aims to discuss fixed-time tracking control problem for a nonholonomic wheeled mobile robot based on visual servoing. At first, by making use of the pinhole camera model, the robot system model with uncalibrated camera parameters is given. Then, the tracking error system between the mobile robot and desired trajectory is proposed. Thirdly, on the basis of fixed-time control theory and Lyapunov stability analysis, fixed-time tracking control laws are proposed for the robot, which can make the robot achieve the desired value in a fixed time. Simulation results are given at the end.

Gantry cranes, which have attracted extensive attention, are mostly simplified as nonlinear single pendulum systems without load hoisting/lowering. However, in fact, due to the existence of the hook, gantry cranes produce double pendulum swing. With an extra underactuated degree of freedom, the anti-swing control of the double pendulum gantry cranes becomes more difficult than that of single pendulum gantry cranes, especially when load hoisting/lowering is considered simultaneously. Moreover, large swings, which lead to problems such as inaccurate positioning and low transportation efficiency, may be caused by double pendulum gantry cranes with load hoisting/lowering. In this paper, a nonlinear coupled tracking anti-swing controller is proposed to solve these problems. In this controller, a smooth tracking trajectory is introduced to ensure the stable start and run of the trolley, and a coupled signal is constructed to eliminate the residual swing angles of gantry crane system. The system stability is analyzed by using Lyapunov method and Barbarat theorem. Theoretical derivation, simulation and experimental results show that the proposed controller has excellent control performance, specifically, not only does it ensure accurate positioning of the load, but also it suppresses and eliminates the hook/load swing angle effectively. The proposed controller can still achieve good control effects and has strong robustness under the condition of changing the load mass, trolley target displacement, system initial swing angles and adding external disturbance.

This paper presents a finite-time command-filtered approximation-free attitude tracking control for rigid spacecraft. A novel finite-time prescribed performance function (FTPPF) is first constructed to ensure that the attitude tracking errors converge to the predefined region in finite time. Then, a finite-time error compensation mechanism is constructed and incorporated into the backstepping control design, such that the differentiation of virtual control signals in recursive steps can be avoided to overcome the singularity issue. Compared with most of approximation-based attitude control methods, less computational burden and lower complexity are guaranteed by the proposed approximation-free control scheme due to the avoidance of using any function approximations. Simulations are given to illustrate the efficiency of the proposed method.

A thunderstorm tracking algorithm is proposed to nowcast the possibility of lightning activity over an area of concern by using the total lightning data and neighborhood technique. The lightning radiation sources observed from the Beijing Lightning Network (BLNET) were used to obtain information about the thunderstorm cells, which are significantly valuable in real-time. The boundaries of thunderstorm cells were obtained through the neighborhood technique. After smoothing, these boundaries were used to track the movement of thunderstorms and then extrapolated to nowcast the lightning approaching in an area of concern. The algorithm can deliver creditable results prior to a thunderstorm arriving at the area of concern, with accuracies of 63%, 80%, and 91% for lead times of 30, 15, and 5 minutes, respectively. The real-time observations of total lightning appear to be significant for thunderstorm tracking and lightning nowcasting, as total lightning tracking could help to fill the observational gaps in radar reflectivity due to the attenuation by hills or other obstacles. The lightning data used in the algorithm performs well in tracking the active thunderstorm cells associated with lightning activities.

The existing pedestrian tracking applications are challenging to balance real-time performance and accuracy. We propose a detection–tracking–correction strategy based on the improved single-shot multi-box detector (SSD), Deep-SORT, and the improved multi-stage object detection architecture (Cascade-R-CNN), which takes both real-time performance and accuracy into consideration. For the detection mechanism, the SSD network is fast and efficient, but the disadvantage of the SSD network is relatively low accuracy. Therefore, the tricks such as cross-entropy loss function, deconvolution, and non-maximum suppression are introduced to improve the SSD network. Then, the improved SSD network is used as the central pedestrian detector to ensure real-time performance. For the tracking mechanism, the Deep-SORT is used to improve the mismatch between tracking and detection. For the correction mechanism, the improved Cascade R-CNN (introducing deformable convolution and group normalization) is used as the reference network to correct the detection errors. The experiment on the data set OTB-100 shows that the proposed strategy has good stability and adaptability in various complex scenes, and the conditions of missed detection and false detection are significantly reduced.

As augmented reality technologies develop, real-time interactions between objects present in the real world and virtual space are required. Generally, recognition and location estimation in augmented reality are carried out using tracking techniques, typically markers. However, using markers creates spatial constraints in simultaneous tracking of space and objects. Therefore, we propose a system that enables camera tracking in the real world and visualizes virtual visual information through the recognition and positioning of objects. We scanned the space using an RGB-D camera. A three-dimensional (3D) dense point cloud map is created using point clouds generated through video images. Among the generated point cloud information, objects are detected and retrieved based on the pre-learned data. Finally, using the predicted pose of the detected objects, other information may be augmented. Our system estimates object recognition and 3D pose based on simple camera information, enabling the viewing of virtual visual information based on object location.

The use of low-cost immersive virtual reality systems is rapidly expanding. Several studies started to analyse the accuracy of virtual reality tracking systems, but they did not consider in depth the effects of external interferences in the working area. In line with that, this study aimed at exploring the static-positional accuracy and the robustness to occlusions inside the capture volume of the SteamVR (1.0) tracking system. To do so, we ran 3 different tests in which we acquired the position of HTC Vive PRO Trackers (2018 version) on specific points of a grid drawn on the floor, in regular tracking conditions and with partial and total occlusions. The tracking system showed a high inter- and intra-rater reliability and detected a tilted surface with respect to the floor plane. Every acquisition was characterised by an initial random offset. We estimated an average accuracy of 0.5 ± 0.2 cm across the entire grid (XY-plane), noticing that the central points were more accurate (0.4 ± 0.1 cm) than the outer ones (0.6 ± 0.1 cm). For the Z-axis, the measurements showed greater variability and the accuracy was equal to 1.7 ± 1.2 cm. Occlusion response was tested using nonparametric Bland–Altman statistics, which highlighted the robustness of the tracking system. In conclusion, our results promote the SteamVR system for static measures in the clinical field. The computed error can be considered clinically irrelevant for exercises aimed at the rehabilitation of functional movements, whose several motor outcomes are generally measured on the scale of metres.

The installation of vibration-based structural health-monitoring systems relying in the study of natural frequencies to assess the condition of civil engineering structures is becoming more and more common. The extraction of these monitoring features can be achieved through automated operational modal analysis, combining output-only identification methods with cluster analyses, and comparing the identified modal properties with references, in a process known as modal tracking. However, changing environmental and operational conditions affect the dynamic behaviour of structures, disturbing the process of modal tracking, which may lead to the loss of important information and to misidentifications. In this context, this paper proposes a new methodology for modal tracking within the scope of automated operational modal analysis, especially prepared for scenarios with strong external influence on modal properties. A concrete arch dam with large variations of some natural frequencies is used as case study, and the proposed methodology is compared to a standard procedure using a quite unique monitoring data set continuously collected for 3 years.

With the growing number of satellite launches, the tracking of resident space objects has become increasingly more important. The application of spacecraft formation flying for complex mission concepts can afford the opportunity for higher resolution imaging and precise tracking of targets in orbit. In-orbit target tracking using formations has many advantages over ground-based approaches, from improving space situational awareness to clearly detecting spacecraft exterior damage. This paper investigates the advantages of in-orbit target tracking considering three different strategies, namely flyby, natural formation, and a novel all-facet-imaging formation. For each strategy, a scenario is developed where an observer spacecraft with an imaging payload attempts to observe a target object while flying in the vicinity of the target. The attitude and orbital dynamics of the observer spacecraft are formulated for each scenario, and two guidance laws for controlling the observer’s orbit and attitude during the tracking are formulated. To evaluate the viability of the tracking strategies, case studies are performed, which consider disturbance torques and perturbative forces. The effectiveness of each approach with respect to tracking performance and required propellant is discussed.

The micro-hole array machined by micro-electrical discharge machining (micro-EDM) has been widely used in aerospace, electronic communication, and other industrial fields. However, due to the unavoidable electrode wear and the unstable discharge state during the electrode breakthrough, the electrode breakthrough detection is usually inaccurate, and the servo tracking performance is also poor, which has adverse effect on the machining efficiency and quality of micro-hole array. In the present study, a three-dimensional flow field model was established, and the gap flow field before and after electrode breakthrough based on the model was analyzed. Simulation results showed that the flow field pattern of dielectric was changed and the flow velocities in bottom gap increased in general after electrode broke through the workpiece, which enabled to analyze the influence of the variation in flushing mechanism of debris on the discharge state in breakthrough stage to some extent. A scheme was proposed for online detecting the breakthrough by monitoring the variations of the gap voltage and the electrode displacement, and the verification experiment was carried out. The proposed scheme has been shown to guarantee the electrode breakthrough, while in the same time controlling the distance of electrode feeding past the micro-hole outlet not exceeding 25 μm. In addition, a segmented servo strategy was proposed to ensure the machining performance of micro-hole array, and it was experimentally tested. This strategy could reduce the average machining time of micro-hole array by about 36% and ensure the machining quality of micro-hole outlet.

Friction stir extrusion (FSE) is a novel solid-phase processing technique that consolidates and extrudes metal powders, flakes, chips, or billets into high-performance parts by plastic deformation, which has the potential to save substantial processing time and energy. Currently, most studies on FSE are experimental and only a few numerical models have been developed to explain and predict the complex physics of the process. In this work, a meshfree simulation framework based on smoothed particle hydrodynamics (SPH) was developed for FSE. Unlike traditional grid-based methods, SPH is a Lagrangian particle-based method that can handle severe material deformations, capture moving interfaces and surfaces, and monitor the field variable histories explicitly without complicated tracking schemes. These aspects of SPH make it attractive for the FSE process, where in situ evolution of field variables is difficult to observe experimentally. To this end, a 3-D, fully thermomechanically coupled SPH model was developed to simulate the FSE of aluminum wires. The developed model was thoroughly validated by comparing the numerically predicted material flow, strain, temperature history, and extrusion force with experimental results for a certain set of process parameters. The validated SPH model can serve as an effective tool to predict and better understand the extreme thermomechanical conditions during the FSE process.

This paper investigates the problem of fixed-time tracking control for a class of high-order nonlinear systems with matched disturbances. A novel continuous fixed-time sliding mode disturbance observer is first proposed to accurately estimate the external disturbances. Then, a new integral high-order sliding mode (IHOSM) surface is proposed in the sense of fixed-time stability by the bi-limit homogeneous method. Subsequently, utilizing the disturbance estimation information, an IHOSM-based fixed-time control scheme is proposed which can enforce the closed-loop control system reach the real sliding mode surface. Meanwhile, it is applied to an error dynamic system of a wheeled mobile robot to achieve fast accurate trajectory tracking. Finally, the comparative experiment results demonstrate the effectiveness and superiority of the proposed control approach.

Visual tracking is an important research topic in the field of computer vision. The Siamese network tracker based on the region proposal network has achieved promising tracking results in terms of speed and accuracy. However, for fast-moving objects, the structure of the tracking system mainly focuses on information regarding the object appearance, ignoring information related to movement and change at any moment. The original 2D convolutional neural network cannot extract the spatiotemporal information of tracking object and cannot pay attention to the features of tracking object. In this research, a new tracking method is proposed that can extract the spatiotemporal features of tracking objects by constructing a 3D convolutional neural network and integrating the cascade attention mechanism and distinguish similar objects by background suppression and highlighting techniques. To verify the effectiveness of the proposed tracker (STASiamRPN), experiments on the OTB2015, GOT-10K and UAV123 benchmark datasets demonstrated that the proposed tracker was highly comparable to other state-of-the-art methods.

A key function of multitarget tracking (MTT) is the state estimation of unknown targets. In this paper, we propose a new state estimator, the probabilistic data association finite impulse response filter (PDAFIRF), which is designed to overcome the drawbacks of the existing state estimators used for MTT. Because the existing state estimation methods used for MTT employ all past measurements, they may exhibit poor performance because of the accumulated errors caused by modeling uncertainties and numerical errors. To overcome the weaknesses of the existing methods, the proposed PDAFIRF uses recent finite measurements, and therefore, can prevent accumulated errors. The proposed PDAFIRF requires recent finite measurements from lidar and stereo vision and processes data in a probabilistic manner. Owing to its FIR-type structure, the PDAFIRF overcomes the structural and conditional defects of the existing stochastic filters. An MTT algorithm employing the PDAFIRF and lidar and stereo vision data is developed for multi-object tracking and target information estimation with high accuracy. The fusion of lidar and stereo vision sensor data is provided for the PDAFIRF. For verifying the high accuracy of the PDAFIRF-based MTT, a simulation of tracking eight objects under fast-varying conditions is conducted. An experimental test is performed using lidar and stereo vision for tracking three pedestrians. It is demonstrated that the proposed PDAFIRF-based MTT considerably enhances the tracking performance compared to the existing Kalman filter-, unscented Kalman filter-, and particle filter-based algorithms.

To make a better discrimination between target and false alarm, amplitude information (AI) of radar signal has been incorporated into many target tracking algorithms, such as probabilistic data association, multiple hypothesis tracking and probability density hypothesis (PHD). In this paper, we present the cubature integration based Gaussian mixture implementation of the PHD filter with AI, namely the GMPHD-AI filter. Since the utilization of multiple sensors is more effective than using only one sensor, we also extend the GMPHD-AI filter to the multi-sensor application where the iterated-corrector scheme is exploited for multi-sensor information fusion. The effectiveness of the presented method is demonstrated by a multiple targets tracking scenario.

This paper deals with a model-free closed-loop fuzzy logic controller development to solve the speed tracking problem for a permanent magnet synchronous motor. The design procedure takes advantage of the fuzzy Lyapunov synthesis approach supporting the stability of the system. In contrast to traditional design control techniques, this solution does not require an a priori knowledge of the plant’s mathematical model. On the other hand, with an advantageous practical meaning, the fuzzy logic controller can be estimated through qualitative knowledge about the physical plant behavior. Lyapunov theory takes action allowing find rules that ensure stability instead of the traditional fuzzy logic controller where formal control strategies are involved. Additionally, the proposal is compared against well-known PI-controller under the same dynamical conditions, obtaining a more efficient control in terms of energy consumption without losing performance. The designed fuzzy controller performance is validated with simulations and experiments, verifying that the designed fuzzy controller can accurately solve the tracking problem on permanent magnet synchronous motors.

In this paper, we proposed a multi-view system for 3D human skeleton tracking based on multi-cue fusion. Multiple Kinect version 2 cameras are applied to build up a low-cost system. Though Kinect cameras can detect 3D skeleton from their depth sensors, some challenges of skeleton extraction still exist, such as left–right confusion and severe self-occlusion. Moreover, human skeleton tracking systems often have difficulty in dealing with lost tracking. These challenges make robust 3D skeleton tracking nontrivial. To address these challenges in a unified framework, we first correct the skeleton's left–right ambiguity by referring to the human joints extracted by OpenPose. Unlike Kinect, and OpenPose extracts target joints by learning-based image analysis to differentiate a person's front side and backside. With help from 2D images, we can correct the left–right skeleton confusion. On the other hand, we find that self-occlusion severely degrades Kinect joint detection owing to incorrect joint depth estimation. To alleviate the problem, we reconstruct a reference 3D skeleton by back-projecting the corresponding 2D OpenPose joints from multiple cameras. The reconstructed joints are less sensitive to occlusion and can be served as 3D anchors for skeleton fusion. Finally, we introduce inter-joint constraints into our probabilistic skeleton tracking framework to trace all joints simultaneously. Unlike conventional methods that treat each joint individually, neighboring joints are utilized to position each other. In this way, when joints are missing due to occlusion, the inter-joint constraints can ensure the skeleton consistency and preserve the length between neighboring joints. In the end, we evaluate our method with five challenging actions by building a real-time demo system. It shows that the system can track skeletons stably without error propagation and vibration. The experimental results also reveal that the average localization error is smaller than that of conventional methods.

Continuously detecting traffic signs in a video sequence is necessary for autonomous or assisted driving scenarios, since a vehicle needs the information from the signs to facilitate navigation. Single-image based traffic sign detector may fail in many cases, when the car moves fast on the road, resulting in motion blur, partial occlusion, and abrupt environmental change. In this paper, we propose an effective methodology, called detection-by-tracking, for robust traffic sign detection in videos, so as to improve the detection performance beyond a basic object detector. We explore the temporal cues among frames to help with the proposal reasoning for further regression. The correlations of spatial location and appearance similarity for the same sign in adjacent frames are considered in our approach. Experimental results show that the proposed detection-by-tracking mechanism is helpful, with improved detection performance to a large extent. Moreover, the idea of the detection-by-tracking can also be generalized to other scenarios for object detection tasks in videos.

Camshift algorithm tracking is susceptible to interference when a tracking object is occluded or when its hue is similar to the background. An improved Camshift object-tracking algorithm combining AKAZE (Accelerated-KAZE) feature matching and Kalman filtering is proposed. First, the video channel is converted for processing. Second, AKAZE is used to match the object feature points and Kalman filtering is used to predict the next position. Then different scenes are judged by the threshold and the Camshift and Kalman tracking algorithms are used for object tracking, respectively. Finally, the improved Camshift algorithm is used to test the moving object in a variety of situations and compared with the traditional Camshift algorithm and the Kalman filter improved Camshift algorithm. Experimental results show that the improved joint tracking algorithm can continue tracking under full occlusion. The effective frame rate of recognition is increased by about 20%, and the single-frame image processing time is less than 35 ms, which can meet the real-time tracking requirements.

Real-time tumor tracking radiotherapy (RTRT) systems typically use fiducial markers implanted near the tumor to track the target using X-ray fluoroscopy. Template pattern matching, used in tracking, is often used to automatically localize the fiducial markers. In radiotherapy of the liver, the thickness of the body that can recognize the fiducial markers must be clinically assessed. The purpose of this study was to quantify the recognition of fiducial markers according to body thickness in stereotactic body radiotherapy of the liver using clinical images obtained using SyncTraX FX4. The recognition scores of fiducial markers were examined in relation to water equivalent length (WEL), tube current, and each flat panel detector. The relationship between the contrast ratio of the fiducial marker and the background and the WEL was also investigated. The average recognition score was found to be less than 20 when the WEL was greater than 25 cm. The probability of successful tracking of image recognition was mostly smaller than 0.8 when the WEL was over 30 cm. The relationship between WEL and tube current did not significantly differ between 100 and 140 mA, but there was a significant difference (p

Multiple Object Tracking (MOT) has attracted increasing interests in recent years, which plays a significant role in video analysis. MOT aims to track the specific targets as whole trajectories and locate the positions of the trajectory at different times. These trajectories are usually applied in Action Recognition, Anomaly Detection, Crowd Analysis and Multiple-Camera Tracking, etc. However, existing methods are still a challenge in complex scene. Generating false (impure, incomplete) tracklets directly affects the performance of subsequent tasks. Therefore, we propose a novel architecture, Siamese Bi-directional GRU, to construct Cleaving Network and Re-connection Network as trajectory post-processing. Cleaving Network is able to split the impure tracklets as several pure sub-tracklets, and Re-connection Network aims to re-connect the tracklets which belong to same person as whole trajectory. In addition, our methods are extended to Multiple-Camera Tracking, however, current methods rarely consider the spatial-temporal constraint, which increases redundant trajectory matching. Therefore, we present Position Projection Network (PPN) to convert trajectory position from local camera-coordinate to global world-coodrinate, which provides adequate and accurate temporal-spatial information for trajectory association. The proposed technique is evaluated over two widely used datasets MOT16 and Duke-MTMCT, and experiments demonstrate its superior effectiveness as compared with the state-of-the-arts.

The accuracy of the CyberKnife Synchrony Respiratory Tracking System is dependent on the breathing pattern of a patient. Therefore, the tracking error in each patient must be determined. Support vector regression (SVR) can be used to easily identify the tracking error in each patient. This study aimed to develop a system with SVR that can predict tracking error according to a patient's respiratory waveform.Datasets of the respiratory waveforms of 93 patients were obtained. The feature variables were variation in respiration amplitude, tumor velocity, and phase shift between tumor and the chest wall, and the target variable was tracking error. A learning model was evaluated with tenfold cross-validation. We documented the difference between the predicted and actual tracking errors and assessed the correlation coefficient and coefficient of determination.The average difference and maximum difference between the actual and predicted tracking errors were 0.57 ± 0.63 mm and 2.1 mm, respectively. The correlation coefficient and coefficient of determination were 0.86 and 0.74, respectively.We developed a system for obtaining tracking error by using SVR. The accuracy of such a system is clinically useful. Moreover, the system can easily evaluate tracking error. We developed a system that can be used to predict the tracking error of SRTS in the CyberKnife Robotic Radiosurgery System using machine learning. The feature variables were the breathing parameters, and the target variable was the tracking error. We used support vector regression algorithm.

An analytic solution to the optimal formation problem of an unmanned aerial vehicle (UAV) swarm is newly proposed for maximizing the perfomrance of passive target tracking. Most previous techniques could not handle this problem efficiently because they have had difficulties in defining the performance index of a nonlinear target tracking filter in closed form. To overcome this limitation, the passive target tracking problem is investigated within the framework of the linear non-conservative robust Kalman filter (NCRKF) theory. Accommodating the merit of the suggested linear target tracking filter structure, its performance measure can be analytically expressed in terms of the UAV formation as well as the sensor accuracy. Therefore, it is easy to determine the optimal UAV formation by maximizing the target tracking performance in the worst case. In addition, our approach is very practical because it considers the estimation error characteristics of the actual passive target tracking filter and the communication range among UAVs in determining the optimal UAV formation. Through the simulations, the effectiveness of the proposed scheme is validated.

This paper develops a robust regressor-free controller for n-link robots with state constraints. We use the function approximation technique to represent the uncertain robot dynamics. Our controller, which uses a state-dependent barrier Lyapunov function, prevents the states from violating their constraints by guaranteeing uniform ultimate boundedness of the closed-loop dynamics via a fixed control structure and a continuous switching law. In contrast to the drawbacks of controllers developed via the widely used error-dependent barrier Lyapunov functions, our controller overcomes the need to impose constraints on the initial tracking errors and the reference trajectories. When the reference trajectories violate the state constraints, our controller is capable of maintaining states within their constraints. We show that it is easy to switch between the constrained and unconstrained versions of our controller. Our controller requires few tuning parameters and is robust to wide ranges of uncertainties. Simulation and real-time experimental tests validate the practicality of our proposed controller.

This paper introduces an anti-slip controller for a quarter vehicle traction control system. By applying the direct Lyapunov method, the single-loop anti-slip control is formulated. However, in the presence of disturbances and parameter uncertainties such as the tire-road interaction and aerodynamic forces, the standard direct Lyapunov method cannot achieve control requirements. In order to tackle the problem, high gain observers are used to improve the system performance. The observer is constructed to estimate driving force, total driving resistance, and disturbances acting on the system. Besides, vehicle speed detection is generally needed in a vehicle control system. However, it is challenging to measure vehicle speed directly. In this paper, the controller and observers are proposed without detecting the vehicle speed. Subsequently, stability of the whole system is proven. The effectiveness and feasibility of the proposed method are demonstrated by simulation results.

A smart transportation system relies on connected environments and cloud systems for ease of operation and assisted routing. Smart vehicles understand the environment through multi-sensor, network, and pervasive systems for gaining useful information. The problem arises with the absence of useful information in explicit scenarios where heterogeneous information becomes mandatory. This article aims to improve transportation support's effectiveness using a discrete behavior information fusion (DBIF) based on the deep learning technique by considering the contradiction in information availability. This proposed technique observes the vehicles' behavior and response to the scene throughout the route displacement. The deep learning model achieves greater accuracy in target detection and classification. The learning output is the independent fusion of behavior (response output) and information (sensed). This sensed information is useful in categorizing further deviations and stipulations for the progressive displacements. The stipulated information and its deviations are recurrently categorized using support vector machine learning. The information provides accurate positioning and tracking of smart vehicles by reducing approximation errors and complexity. The simulation results illustrate the proposed technique's efficiency by improving the accuracy of 92.078% and fusion rate of 0.9741 and reducing error of 0.0662, complexity of 0.0717, and fusion time of 0.9938 compared to existing methods.

Camera calibration is essential for calculating camera parameters, and the parameters should be constant whenever the position tracker needs to be recalibrated. Manual camera calibration obtains a checkerboard image by tilting or positioning the checkerboard by hand. However, it makes difficult to obtain the same checkerboard images in every calibration, resulting in different camera parameters. Therefore, this study proposes robotic camera calibration to produce the constant camera parameters whenever the tracker needs to be calibrated. The robot arm moves the checkerboard according to the programmed positions and orientations so that the position tracker can obtain the same checkerboard image every calibration. Experiments were conducted to compare results of the manual and robotic camera calibrations. First, we compared the standard deviations of the intrinsic parameters and extrinsic parameters; the results show that robotic camera calibration produced the parameters with smaller standard deviations. We also compared the tracking precision of the cubic marker using manual and robotic camera calibration. As a result, the standard deviations of the tracking precision of cubic marker by the manual and the robotic camera calibration were 1.65 mm and only 0.35 mm, respectively. It reveals that the robotic camera calibration enables more precise and consistent tracking than the manual calibration.

A T-S fuzzy model identification method based on physical membership function is proposed for maneuvering target tracking. T-S fuzzy model is a good tool for fitting complex and nonlinear systems conducted by separating inputs–outputs spaces of systems and identifying corresponding parameters. Usually membership degrees play important roles in fusing local fuzzy models for reflecting nonlinear property of T-S fuzzy model, however, usual membership degrees are obtained by Gaussian function, which not only lacks interpretability and meanings but also is complex to be used. In this paper, a physical membership function with interpretability and physical meanings is proposed. To identify T-S fuzzy model based on the proposed method, first, a hyper-planed FSC algorithm as the separating method is utilized. Then UKF is used to identify consequent parameters. Finally, the proposed physical membership function is used to fuse local models and estimate final states. We apply the proposed T-S fuzzy model algorithms to maneuvering target tracking, and comparisons with several classical methods on both simulated data and real data demonstrate effectiveness and advantages of the proposed methods in tracking accuracy.

In recent years, object tracking based on Siamese network has attracted the attention by taking advantage of both speed and accuracy. However, when similar object interference, severe occlusion and other factors cause temporary tracking failure, a restricted search window makes Siamese network difficult to retrieve the object again, and results in unrecoverable tracking failures. In this paper, a general and efficient two-way verification tracking failure detection method using parallel correlation filtering and Siamese network is proposed, which can detect the tracking failure and retrieve the object again. Firstly, we construct parallel Siamese network and correlation filter tracking network, and get tracking results respectively. Secondly, we make a preliminary judgment on the reliability of the tracking results based on the overlap ratio of the tracking results. Finally, based on two-way verification, which tracker failed to track is finally determined, and the search window of Siamese network is optimized to retrieve the object again. We comparisons with state-of-the art trackers on benchmark datasets: OTB100, VOT2016, VOT2018, VOT2019 and NFS. The results show that the method we proposed can detect the tracking failure and retrieve the object again, thus improving the accuracy of object tracking.

Multiple speaker localization and tracking in noisy and reverberant environments is a challenging task. A distributed multiple speaker tracking method based on unscented particle filter and data association is proposed in microphone array networks. First, the generalized cross correlation (GCC) function is estimated at each node and the time difference of arrival (TDOA) is viewed as observation for tracking. Considering the ambiguity of the observations caused by the interference of other acoustic sources and multipath, the data association technique is used at each node to associate the available observations with each speaker. Then according to the characteristics of GCC function, the association probability of valid observations for each speaker is calculated by combining the amplitude of the valid observations with the statistical distance. After that, the local state of speakers is obtained at each node based on unscented particle filter. Finally, according to the reliability of the local state of each node, a dynamic weighted consensus fusion algorithm is presented to approximate the global state estimation and obtain good multiple speakers tracking performance. The proposed method can track multiple speakers under reverberant and noisy environments in a distributed manner and is scalable and robust against node failure in DMA. Simulation results verify the effectiveness of the proposed method.

This paper considers the adaptive finite-time tracking control problem for a class of switched nonlinear systems. Based on neural networks, the influence of unknown nonlinear functions is offset. A performance function is proposed to guarantee the transient performance of the tracking error. Moreover, a novel-type Lyapunov function is constructed that relaxes the constraints on virtual control functions. It is shown by Lyapunov stability analysis that the finite-time convergence and zero-error tracking performance can be ensured simultaneously. Finally, the effectiveness of the proposed method is verified by a simulation example.

This paper presents the systematic design and testing of a novel two-axial flexure-based vibration stage. The testing principle of transverse sensitivity is first established based on the conceptual level and principle level. According to three kinds of flexure-based kinematic chains, the configuration of this stage is then determined. Subsequently, the design based on increasing the degree of segmentation is discussed. Finally, an experimental platform is set up. The repeated tracking performance and measurement uncertainty of transverse sensitivity are tested. The results indicate that it has high repeated tracking accuracy. Average transverse sensitivities for four different cycles at 80, 120, and 140 Hz are 2.487%, 2.367% and 2.281%. The total average is about 2.378%, while the total measurement uncertainty is 0.149% at these different frequencies.

In this study, a multi-object tracking (MOT) scheme based on a light detection and ranging sensor was proposed to overcome imprecise velocity observations in object occlusion scenarios. By applying real-time velocity estimation, a modified unscented Kalman filter (UKF) was proposed for the state estimation of a target object. The proposed method can reduce the calculation cost by obviating unscented transformations. Additionally, combined with the advantages of a two-reference-point selection scheme based on a center point and a corner point, a reference point switching approach was introduced to improve tracking accuracy and consistency. The state estimation capability of the proposed UKF was verified by comparing it with the standard UKF in single-target tracking simulations. Moreover, the performance of the proposed MOT system was evaluated using real traffic datasets.

This paper studies the distributed optimization problem over an undirected connected graph subject to digital communications with a finite data rate, where each agent holds a strongly convex and smooth cost function. The agents need to cooperatively minimize the average of all agents’ cost functions. Each agent builds an encoder/decoder pair that produces transmitted messages to its neighbors with a finite-level uniform quantizer, and recovers its neighbors’ states by a recursive decoder with received quantized signals. Combining the adaptive encoder/decoder scheme with the gradient tracking method, the authors propose a distributed quantized algorithm. The authors prove that the optimization can be achieved at a linear rate, even when agents communicate at 1-bit data rate. Numerical examples are also conducted to illustrate theoretical results.

The advance of visual tracking has provided unmanned aerial vehicle (UAV) with the intriguing capability for various practical applications. With promising performance and efficiency, discriminative correlation filter (DCF)-based trackers have drawn significant attention and undergone remarkable progress. However, the boundary effect and filter degradation remain two intractable problems. In this work, we propose a novel Adaptive Spatio-Temporal Regularized Correlation Filters (ASTR-CF) model to address the two problems. The ASTR-CF model simultaneously optimizes the spatial and temporal regularization weights adaptively, and it is optimized by the alternating direction method of multipliers (ADMM) effectively. Extensive experiments on 4 UAV tracking benchmarks have proven the superiority of the proposed ASTR-CF compared with more than 30 state-of-the-art trackers in terms of accuracy and speed.

Multi-object tracking is a vital problem as many applications require better tracking approaches. Although learning-based detectors are becoming extremely powerful, there are few tracking methods designed to work with them in real time. We explored an efficient flexible online vehicle tracking-by-detection framework suitable for real-virtual mapping systems, which combines a non-recursive temporal window search with delayed output and produces stable trajectories despite noisy detection responses. Its computation speed meets the real-time requirements, whereas its performance is comparable with that of state-of-the-art online trackers on the DETRAC dataset. The trajectories from our approach also contain the target class and color information important for virtual vehicle motion reconstruction.

Automated excavation is hard to achieve due to several inherent problems such as resistive force acting against the bucket, non-homogenous dynamics of various excavation media, and nonlinearities of the excavator’s hydraulics system. To deal with this issue, this paper provides an integrative control strategy for successful autonomous excavation that considers the mutually associated factors, i.e., position, contour, and force control. For the position tracking, a non-linear PI controller was designed to track the position of individual actuators of the excavator and thereby control the bucket tip’s position. In addition, the contour control technique was applied to achieve an optimal excavation path to minimize contour errors. Finally, to compensate for the ground resistive force during digging tasks, a force impedance controller was designed along with the time-delayed control that reduces the effect of dynamic uncertainties. Experimental results with a modified mini-wheeled excavator show that the developed integrative tracking control strategy can provide a comprehensive solution to improving the tracking performance for autonomous excavation that can simultaneously deal with the critical components of position, contour, and force control.

This article introduces the basic element to build an augmented reality system allowing to insert a 2D object in a real 3D scene in real time. The first step consists in locating the place where to insert the object using an abstract marker, this marker is a rectangle that surrounds the minimum area of the object's contour detected through its color. This rectangle is a plane surface that provides the position of its four points in the images acquired in real time which allows to have a real time tracking of the detected object. Planar homography describes exactly the relationship between the key points if the scene is flat and only requires four key points to produce an exact solution of the camera position to align a 2D virtual object in a 3D real scene.

Electromigration (EM) is a severe reliability issue in power grid networks. The via array possesses special EM characteristics and suffers from Joule heating and current crowding, closely related to EM violations. In this study, a power grid EM analysis method was developed to solve temperature variation effects for the via array EM. The new method is based on the temperature-aware EM model, which considers the effects of self-heating and thermal coupling of interconnected lines in a power grid. According to the model, the proposed methodology introduces a locality-driven strategy and current tracking to perform full-chip EM assessment for multilayered power grids. The results show that temperature due to Joule heating indeed has significant impacts on the via EM failure. The results further demonstrate that the proposed method might reasonably improve efficiency while ensuring the accuracy of the analysis.

Under the low speed condition, a method of real-time tracking and estimation of rotor position based on PLL technology is proposed, which is used to solve the control system detection accuracy problem of permanent magnet synchronous motor (PMSM) for electric vehicles. The control principles of high frequency signal fluctuation are analyzed, and the mathematical model of three phases PMSM under rotor estimated synchronous rotating reference frame is established. The basic principles of phase locked loop (PLL) are analyzed. Based on phase locked loop, a rotor position estimation method is designed and analyzed. Finally, simulation model of sensorless control system is set up, and the simulation experiment is carried out. The simulation experiment results show that the sensorless control based on PLL can obtain the accurate rotor positions and the excellent control ability. Therefore, the rotor positions estimation method based on PLL is an ideal method for the sensorless control of electric vehicle drive motor, which can provide theoretical and technical support for improving the control precision of PMSM and quality of electric vehicles.

Intracellular organelles play an essential role in regulating the biochemistry of various processes and events in eukaryotic cells. Understanding and quantifying their morphological dynamics should allow for their deeper analysis. However, the development of organelle and membrane tracking algorithms is challenging, especially in the case of studying the regulation of Golgi-derived membrane tubules, which show varied sizes and shapes, continuous fission and fusion events, and different directions of movement. Here we have sought to establish a tracking system for tubular subcellular structures, to characterize morphological changes in membrane fission and fusion events. The development of a tracking algorithm consists of two methods of particle linking: the k-nearest neighbor method applied to successive images with a high temporal resolution, and a modification method named fixed-lag interval smoothing Kalman filtering, for situations of temporary particle disappearance and forward and backward movement. The system shows excellent efficiency for tracking. One key advantage of this system is that it not only provides track mapping of tubules for fast and convenient interpretation, but also displays tubule labeling and the type of fission and fusion, with corresponding changes of events displayed in the track mapping view. Moreover, tracking measurements of event type and growth rate in length, motility, velocity, and diffusion features are calculated and saved offline use for further analysis. This prototype tracking system can be applied to different organelles undergoing fission and fusion events, e.g. Golgi-derived tubules and mitochondria, to help biologists quantify and better understand the mechanisms underlying changes in membrane dynamics.

This paper presents a Linear Quadratic Regulator (LQR) control strategy for a non-linear 3-DOF helicopter system. The optimal tracking approach yields robustness to both structured (parameter variation) and unstructured (disturbance) uncertainties. First, the 3-DOF helicopter system is linearized around an operating point and its state-space model is derived. Then, the LQR controller is designed. Numerical results show the controller’s robustness and high performance under uncertainties. In comparison with other existing control strategies, the control approach proposed in this paper is a lot more simple, hence its comparison with the PID controller, an equally simple control strategy. To better evaluate the execution and the performance of the LQR control strategy, two quantitative tracking performance metrics are presented; (i) the integral of the tracking errors, and (ii) the integral of the control signals of the system. Comparative results show the prevalence of the proposed strategy, as it achieves the highest tracking accuracy with the lowest control exertion.

We study the dynamics of classical localization in a simple, one-dimensional model of a tracking chamber. The emitted particle is represented by a superposition of Gaussian wave packets moving in opposite directions, and the detectors are two spins in fixed, opposite positions with respect to the central emitter. At variance with other similar studies, we give here a phase-space representation of the dynamics in terms of the Wigner matrix of the system. This allows a better visualization of the phenomenon and helps in its interpretation. In particular, we discuss the relationship of the localization process with the properties of entanglement possessed by the system.

Because of the rise of deep learning and neural networks, algorithms based on deep learning have also been developed and subtly applied in daily life. This paper hoped to use neural network-based face recognition with absolute distance estimation and research done with drones to achieve the target tracking effect through the recognition ability of neural networks. The aerial images were used to identify different people or behaviors through which distances could be calculated to establish the foundation for subsequent research and development. According to our experiments, performing the face recognition model and distance estimation met expected results, and it also had a specific accuracy rate from different angles and distances. In summary, identification and distance estimate had better accuracy within the effective length, and we could expect the speed and convenience of identification to be realized on other devices.

How to ensemble the high representative ability for discriminating the target from its background and high adaptive ability to fast appearance changes, while keeping the real-time performance simultaneously is still an open topic in the field of object tracking, especially in the complex urban traffic scenes. To address this issue, motivated by that existing excellent trackers may have their advantages in tackling different kinds of tracking difficulties respectively, we propose a new real-time stage-wise object tracking method that allows different trackers to complement each other and combines their respective advantages. A tracker selection agent is trained to learn the policy of switching to the most appropriate candidate tracker according to the current tracking environment. To capture the dynamics of tracking environment effectively, we consider the tracker selection problem as a Partially Observable Markov Decision Process problem. A lightweight deep neural network with the recurrent unit is designed for learning the optimal policy accurately and rapidly. We also elaborately collected Traffic Scenes Object Tracking Annotated Dataset (TS-OTAD) for demonstrating the effectiveness of our method. Experimental results conducted on TS-OTAD and OTB-100 demonstrate that our method has superior performance than any of the candidate tracker and has a good trade-off between accuracy and efficiency compared with other state-of-the-art methods. Besides, our stage-wise tracking framework is not limited to any specific tracker, and any excellent tracker can be used as the candidate, which provides a new way for boosting object tracking accuracy and efficiency.