scholarly journals Siamese Networks-Based People Tracking Using Template Update for 360-Degree Videos Using EAC Format

Sensors ◽  
2021 ◽  
Vol 21 (5) ◽  
pp. 1682
Author(s):  
Kuan-Chen Tai ◽  
Chih-Wei Tang
Keyword(s):  
High Speed ◽  
Tracking Accuracy ◽  
People Tracking ◽  
Linear Discriminant ◽  
Template Update ◽  
Geometric Deformation ◽  
Plane Projection ◽  
Rich Information ◽  
Score Map ◽  
Siamese Networks

Rich information is provided by 360-degree videos. However, non-uniform geometric deformation caused by sphere-to-plane projection significantly decreases tracking accuracy of existing trackers, and the huge amount of data makes it difficult to achieve real-time tracking. Thus, this paper proposes a Siamese networks-based people tracker using template update for 360-degree equi-angular cubemap (EAC) format videos. Face stitching overcomes the problem of content discontinuity of the EAC format and avoids raising new geometric deformation in stitched images. Fully convolutional Siamese networks enable tracking at high speed. Mostly important, to be robust against combination of non-uniform geometric deformation of the EAC format and partial occlusions caused by zero padding in stitched images, this paper proposes a novel Bayes classifier-based timing detector of template update by referring to the linear discriminant feature and statistics of a score map generated by Siamese networks. Experimental results show that the proposed scheme significantly improves tracking accuracy of the fully convolutional Siamese networks SiamFC on the EAC format with operation beyond the frame acquisition rate. Moreover, the proposed score map-based timing detector of template update outperforms state-of-the-art score map-based timing detectors.

Integrated optimal design of a high-speed planar parallel manipulator

2018 ◽  
Vol 233 (9) ◽  
pp. 2976-2990 ◽  
Author(s):  
Zhengsheng Chen ◽  
Minxiu Kong
Keyword(s):  
Optimal Design ◽  
Parallel Manipulator ◽  
High Speed ◽  
Design Method ◽  
Tracking Error ◽  
Dimensional Synthesis ◽  
Tracking Accuracy ◽  
Nsga Ii ◽  
Parameters Tuning ◽  
Planar Parallel Manipulator

To obtain excellent comprehensive performances of the planar parallel manipulator for the high-speed application, an integrated optimal design method, which integrated dimensional synthesis, motors/reducers selection, and control parameters tuning, is proposed, and the 3RRR parallel manipulator was taken as the example. The kinematic and dynamic performances of condition number, velocity index, acceleration capability, and low-order frequency are taken into accounts for the dimensional synthesis. Then, to match motors/reducers parameters and keep an economical cost, the constraint equations and the parameters library are built, and the cost is chosen as one of the optimization objectives. Also, to get high tracking accuracy, the dynamic forward plus proportional–derivative control scheme is introduced, and the tracking error is chosen as one of the optimization objectives. Hence, the optimization model including dimensional synthesis, motors/reducers selection and controller parameters tuning is established, which is solved by the genetic algorithm II (NSGA-II). The result shows that comprehensive performances can be effectively promoted through the proposed integrated optimal design, and the prototype was constructed according to the Pareto-optimal front.

Trajectory Synthesis and Redundancy Resolution for High Speed Point to Point Motions of Redundant Robot Manipulators

1997 ◽  
Author(s):  
W. Kim ◽  
J. Rastegar
Keyword(s):  
High Speed ◽  
Degrees Of Freedom ◽  
Robot Manipulator ◽  
Robot Manipulators ◽  
Dynamic Performance ◽  
High Harmonic ◽  
Harmonic Excitation ◽  
Redundancy Resolution ◽  
Tracking Accuracy ◽  
Point To Point

Abstract Trajectory synthesis for robot manipulators with redundant kinematic degrees-of-freedom has been studied by numerous investigators. Redundant manipulators are of interest since the redundant degrees-of-freedom can be used to improve the local and global kinematic and dynamic performance of a system. As a robot manipulator is forced to track a given trajectory, the required actuating torques (forces) may excite the natural modes of vibration of the system. Noting that manipulators with revolute joints have nonlinear dynamics, high harmonic excitation torques are generally generated even though such harmonics have been eliminated from the synthesized trajectories and filtered from the drive inputs. In this paper, a redundancy resolution method is developed based on the Trajectory Pattern Method (TPM) to synthesize trajectories such that the actuating torques required to realize them do not contain higher harmonic components with significant amplitudes. With such trajectories, a robot manipulator can operate at higher speeds and achieve higher tracking accuracy with suppressed residual vibration. As an example, optimal trajectories are synthesized for point to point motions of a plane 3R manipulator.

Tracking algorithm of snowboard target in intelligent system

2020 ◽  
pp. 1-9
Author(s):  
Zhipeng Li ◽  
Xiaolan Li ◽  
Ming Shi ◽  
Wenli Song ◽  
Guowei Zhao ◽  
...  
Keyword(s):  
Target Tracking ◽  
High Speed ◽  
Intelligent System ◽  
Mean Shift ◽  
Target Position ◽  
Tracking Algorithm ◽  
Tracking Accuracy ◽  
Mean Shift Algorithm ◽  
Target Track ◽  
The Mean

Snowboarding is a kind of sport that takes snowboarding as a tool, swivels and glides rapidly on the specified slope line, and completes all kinds of difficult actions in the air. Because the sport is in the state of high-speed movement, it is difficult to direct guidance during the sport, which is not conducive to athletes to find problems and correct them, so it is necessary to track the target track of snowboarding. The target tracking algorithm is the main solution to this task, but there are many problems in the existing target tracking algorithm that have not been solved, especially the target tracking accuracy in complex scenes is insufficient. Therefore, based on the advantages of the mean shift algorithm and Kalman algorithm, this paper proposes a better tracking algorithm for snowboard moving targets. In the method designed in this paper, in order to solve the problem, a multi-algorithm fusion target tracking algorithm is proposed. Firstly, the SIFT feature algorithm is used for rough matching to determine the fuzzy position of the target. Then, the good performance of the mean shift algorithm is used to further match the target position and determine the exact position of the target. Finally, the Kalman filtering algorithm is used to further improve the target tracking algorithm to solve the template trajectory prediction under occlusion and achieve the target trajectory tracking algorithm design of snowboarding.

Adaptive Synchronized Control for a Planar Parallel Manipulator: Theory and Experiments

2005 ◽  
Vol 128 (4) ◽  
pp. 976-979 ◽  
Author(s):  
Lu Ren ◽  
James K. Mills ◽  
Dong Sun
Keyword(s):  
Parallel Manipulator ◽  
High Speed ◽  
Control Method ◽  
Parametric Uncertainty ◽  
Synchronization Error ◽  
Tracking Accuracy ◽  
Unknown Parameters ◽  
High Acceleration ◽  
Planar Parallel Manipulator ◽  
Synchronized Control

In this paper, we develop a new control method, termed adaptive synchronized (A-S) control, for improving tracking accuracy of a P-R-R type planar parallel manipulator with parametric uncertainty. The novelty of A-S control, a combination of synchronized control and adaptive control, is in the application of synchronized control to a single parallel manipulator so that tracking accuracy is improved during high-speed, high-acceleration tracking motions. Through treatment of each chain as a submanipulator; the P-R-R manipulator is thus modeled as a multi-robot system comprised of three submanipulators grasping a common payload. Considering the geometry of the platform, these submanipulators are kinematically constrained and move in a synchronous manner. To solve this synchronization control problem, a synchronization error is defined, which represents the coupling effects among the submanipulators. With the employment of this synchronization error, tracking accuracy of the platform is improved. Simultaneously, the estimated unknown parameters converge to their true values through the use of a bounded-gain-forgetting estimator. Experiments conducted on the P-R-R manipulator demonstrate the validity of the approach.

scholarly journals Study on Acc/Dec Algorithm Based on Piecewise Polynomial in CNC

2011 ◽  
Vol 2-3 ◽  
pp. 43-47 ◽  
Author(s):  
Guo Shun Ji ◽  
Zhi Ping Chen ◽  
Ju Yong Zhang ◽  
Wei Liu
Keyword(s):  
High Speed ◽  
Large Scale ◽  
Polynomial Function ◽  
Tracking Accuracy ◽  
Feedrate Planning ◽  
Smooth Motion ◽  
Planning Algorithm ◽  
S Curve ◽  
The Stability ◽  
Motion Profile

In order to improve the stability of feed movement in high speed CNC system, the feedrate planning algorithm based on piece-wise polynomial function was proposed. The flexible transition of feedrate was realized through maintaining linear continuous jerk. The principle of the proposed algorithm was introduced and the method to generate smooth motion profile based on the proposed algorithm was presented. The rapidity, stability and tracking accuracy of the feedrate planning algorithm to linearity, S curve and the proposed one were analyzed. The proposed algorithm is simple and it can be applied in acceleration/deceleration before interpolation in high speed feed movement to improve the stability of it. The proposed algorithm was applied in multi-contour high speed processing and the result indicated that it could improve the stability of large-scale parts motion.

Perfectly Matched Feedback Control and Its Integrated Design for Multiaxis Motion Systems

2004 ◽  
Vol 126 (3) ◽  
pp. 547-557 ◽  
Author(s):  
Syh-Shiuh Yeh ◽  
Pau-Lo Hsu
Keyword(s):  
Feedback Control ◽  
High Speed ◽  
Feedforward Control ◽  
Phase Error ◽  
Integrated Design ◽  
Cnc Machining ◽  
Dynamic Responses ◽  
Tracking Accuracy ◽  
Control Loops ◽  
Contouring Accuracy

For motion systems with multiple axes, the approach of matched direct current gains has been generally adopted to improve contouring accuracy under low-speed operations. To achieve high-speed and high-precision motion in modern manufacturing, a perfectly matched feedback control (PMFBC) design for multiaxis motion systems is proposed in this paper. By applying stable pole-zero cancellation and including complementary zeros for uncancelled zeros for all axes, matched dynamic responses across the whole frequency range for all axes are achieved. Thus, contouring accuracy for multiaxis systems is guaranteed for the basic feedback loops. In real applications, the modeling error is unavoidable and the degradation and limitations of the model-based PMFBC exist. Therefore, a newly designed digital disturbance observer is proposed to be included in the proposed PMFBC structure for each axis to compensate for undesirable nonlinearity and disturbances to maintain the matched dynamics among all axes for the PMFBC design. Furthermore, the feedforward control loops zero phase error tracking controller are employed to reduce tracking errors. Experimental results on a three-axis CNC machining center indicate that both contouring accuracy and tracking accuracy are achieved by applying the present PMFBC design.

scholarly journals Discriminative and Robust Online Learning for Siamese Visual Tracking

2020 ◽  
Vol 34 (07) ◽  
pp. 13017-13024 ◽  
Author(s):  
Jinghao Zhou ◽  
Peng Wang ◽  
Haoyang Sun
Keyword(s):  
Discriminative Learning ◽  
Visual Object ◽  
Specific Context ◽  
Visual Object Tracking ◽  
Robust Learning ◽  
Online Module ◽  
Recent Success ◽  
Template Update ◽  
Update Strategy ◽  
Siamese Networks

The problem of visual object tracking has traditionally been handled by variant tracking paradigms, either learning a model of the object's appearance exclusively online or matching the object with the target in an offline-trained embedding space. Despite the recent success, each method agonizes over its intrinsic constraint. The online-only approaches suffer from a lack of generalization of the model they learn thus are inferior in target regression, while the offline-only approaches (e.g., convolutional siamese trackers) lack the target-specific context information thus are not discriminative enough to handle distractors, and robust enough to deformation. Therefore, we propose an online module with an attention mechanism for offline siamese networks to extract target-specific features under L2 error. We further propose a filter update strategy adaptive to treacherous background noises for discriminative learning, and a template update strategy to handle large target deformations for robust learning. Effectiveness can be validated in the consistent improvement over three siamese baselines: SiamFC, SiamRPN++, and SiamMask. Beyond that, our model based on SiamRPN++ obtains the best results over six popular tracking benchmarks and can operate beyond real-time.

scholarly journals EFFICIENT MULTI-VIEW 3D TRACKING OF ARBITRARY ROCK FRAGMENTS UPON IMPACT

2020 ◽  
Vol XLIII-B2-2020 ◽  
pp. 589-596
Author(s):  
D. E. Guccione ◽  
K. Thoeni ◽  
A. Giacomini ◽  
O. Buzzi ◽  
S. Fityus
Keyword(s):  
3D Reconstruction ◽  
High Speed ◽  
Visual Hull ◽  
Post Processing ◽  
Tracking Accuracy ◽  
3D Tracking ◽  
Rock Fragments ◽  
Fragmentation Test ◽  
Impact Fragmentation ◽  
Improved Performance

Abstract. This paper presents a new methodology to accurately obtain 3D rotational velocities of blocks and fragments. Four high speed cameras are used to capture the scene. An additional two tilted mirrors are used to multiply the number of views. Hence, a total of six different viewing perspectives can be used to track translational and rotational velocities in 3D. The focus in the current work is on the rotational velocities, as tracking of the translation is generally straightforward. A common outline tracking algorithm based on the visual hull is adapted. The visual hull is further meshed using triangular elements to approximate the shape of the object. This 3D reconstruction is then used to track the 3D motion of the object. However, the accuracy of the results strongly depends on the accuracy of the 3D reconstruction which is mainly influenced by the number and position of the available views. In any case, the 3D reconstruction from the visual hull is only an approximation and significant errors can be introduced which influence the tracking accuracy. Hence, an in-house post-processing algorithm based on the knowledge of the real geometry of the object, which can generally be accurately determined after a test, was developed. The improved performance of this new post-processing method is shown by controlled spinning tests. Finally, results of a real example of an impact fragmentation test are discussed.

scholarly journals Non-Contact Measurement of the Surface Displacement of a Slope Based on a Smart Binocular Vision System

Sensors ◽  
2018 ◽  
Vol 18 (9) ◽  
pp. 2890 ◽  
Author(s):  
Leping He ◽  
Jie Tan ◽  
Qijun Hu ◽  
Songsheng He ◽  
Qijie Cai ◽  
...  
Keyword(s):  
Monitoring System ◽  
Template Matching ◽  
High Speed ◽  
Surface Deformation ◽  
Least Squares Method ◽  
Vision System ◽  
Deformation Monitoring ◽  
Tracking Algorithm ◽  
Zernike Moment ◽  
Tracking Accuracy

The paper presents an intelligent real-time slope surface deformation monitoring system based on binocular stereo-vision. To adapt the system to field slope monitoring, a design scheme of concentric marking point is proposed. Techniques including Zernike moment edge extraction, the least squares method, and k-means clustering are used to design a sub-pixel precision localization method for marker images. This study is mostly focused on the tracking accuracy of objects in multi-frame images obtained from a binocular camera. For this purpose, the Upsampled Cross Correlation (UCC) sub-pixel template matching technique is employed to improve the spatial-temporal contextual (STC) target-tracking algorithm. As a result, the tracking accuracy is improved to the sub-pixel level while keeping the STC tracking algorithm at high speed. The performance of the proposed vision monitoring system has been well verified through laboratory tests.

Sign in / Sign up

Export Citation Format

Share Document