Object Tracking on Satellite Videos: A Correlation Filter-Based Tracking Method With Trajectory Correction by Kalman Filter

2019 ◽  
Vol 12 (9) ◽  
pp. 3538-3551 ◽  
Author(s):  
Yujia Guo ◽  
Daiqin Yang ◽  
Zhenzhong Chen
Keyword(s):  
Kalman Filter ◽  
Object Tracking ◽  
Correlation Filter ◽  
Tracking Method ◽  
Trajectory Correction

scholarly journals Spatio-Temporal Context, Correlation Filter and Measurement Estimation Collaboration Based Visual Object Tracking

Sensors ◽  
2021 ◽  
Vol 21 (8) ◽  
pp. 2841
Author(s):  
Khizer Mehmood ◽  
Abdul Jalil ◽  
Ahmad Ali ◽  
Baber Khan ◽  
Maria Murad ◽  
...  
Keyword(s):  
Kalman Filter ◽  
Object Tracking ◽  
Environmental Changes ◽  
Correlation Filter ◽  
Temporal Context ◽  
Visual Object ◽  
Target Model ◽  
Change Of Scale ◽  
Tracking Model ◽  
Spatio Temporal

Despite eminent progress in recent years, various challenges associated with object tracking algorithms such as scale variations, partial or full occlusions, background clutters, illumination variations are still required to be resolved with improved estimation for real-time applications. This paper proposes a robust and fast algorithm for object tracking based on spatio-temporal context (STC). A pyramid representation-based scale correlation filter is incorporated to overcome the STC’s inability on the rapid change of scale of target. It learns appearance induced by variations in the target scale sampled at a different set of scales. During occlusion, most correlation filter trackers start drifting due to the wrong update of samples. To prevent the target model from drift, an occlusion detection and handling mechanism are incorporated. Occlusion is detected from the peak correlation score of the response map. It continuously predicts target location during occlusion and passes it to the STC tracking model. After the successful detection of occlusion, an extended Kalman filter is used for occlusion handling. This decreases the chance of tracking failure as the Kalman filter continuously updates itself and the tracking model. Further improvement to the model is provided by fusion with average peak to correlation energy (APCE) criteria, which automatically update the target model to deal with environmental changes. Extensive calculations on the benchmark datasets indicate the efficacy of the proposed tracking method with state of the art in terms of performance analysis.

Research of ASM Object Tracking Method Combining Kalman Estimation

2014 ◽  
Vol 1049-1050 ◽  
pp. 1595-1598
Author(s):  
Li Guo Zhang ◽  
Mei Jin ◽  
Ju Jin ◽  
Guo Hui Yu
Keyword(s):  
Kalman Filter ◽  
Statistical Model ◽  
Object Tracking ◽  
Experimental Results ◽  
Initial Contour ◽  
Initial State ◽  
Current Frame ◽  
Rigid Object ◽  
Tracking Method ◽  
Actual Contour

ASM is a statistical model applied to match contours of non-rigid object. The actual contour may much different from the initial contour and the result is likely to converge to an error contour. Kalman filter is adopted to track the current frame for the prediction and acts as the initial state of the ASM, and then applies the ASM to correct the contour of the object. Experimental results show that the method proposed in this paper allows the model to converge to the target contour quickly and accurately. It has good stability and robustness.

scholarly journals Learning Local–Global Multiple Correlation Filters for Robust Visual Tracking with Kalman Filter Redetection

Sensors ◽  
2021 ◽  
Vol 21 (4) ◽  
pp. 1129 ◽  
Author(s):  
Jianming Zhang ◽  
Yang Liu ◽  
Hehua Liu ◽  
Jin Wang
Keyword(s):  
Kalman Filter ◽  
Object Tracking ◽  
Correlation Filter ◽  
Appearance Model ◽  
Visual Object ◽  
Multiple Correlation ◽  
Correlation Filters ◽  
Tracking Algorithms ◽  
Global Correlation ◽  
Object Appearance

Visual object tracking is a significant technology for camera-based sensor networks applications. Multilayer convolutional features comprehensively used in correlation filter (CF)-based tracking algorithms have achieved excellent performance. However, there are tracking failures in some challenging situations because ordinary features are not able to well represent the object appearance variations and the correlation filters are updated irrationally. In this paper, we propose a local–global multiple correlation filters (LGCF) tracking algorithm for edge computing systems capturing moving targets, such as vehicles and pedestrians. First, we construct a global correlation filter model with deep convolutional features, and choose horizontal or vertical division according to the aspect ratio to build two local filters with hand-crafted features. Then, we propose a local–global collaborative strategy to exchange information between local and global correlation filters. This strategy can avoid the wrong learning of the object appearance model. Finally, we propose a time-space peak to sidelobe ratio (TSPSR) to evaluate the stability of the current CF. When the estimated results of the current CF are not reliable, the Kalman filter redetection (KFR) model would be enabled to recapture the object. The experimental results show that our presented algorithm achieves better performances on OTB-2013 and OTB-2015 compared with the other latest 12 tracking algorithms. Moreover, our algorithm handles various challenges in object tracking well.

scholarly journals A Kalman Filter-Based Kernelized Correlation Filter Algorithm for Pose Measurement of a Micro-Robot

Micromachines ◽  
2021 ◽  
Vol 12 (7) ◽  
pp. 774
Author(s):  
Heng Zhang ◽  
Hongwu Zhan ◽  
Libin Zhang ◽  
Fang Xu ◽  
Xinbin Ding
Keyword(s):  
Kalman Filter ◽  
Object Tracking ◽  
High Speed ◽  
Correlation Filter ◽  
Moving Targets ◽  
Vision Measurement ◽  
Time Operation ◽  
Micro Robot ◽  
Real Time Operation ◽  
Inter Frame

This paper proposes a moving-target tracking algorithm that measures the pose of a micro-robot with high precision and high speed using the Kalman filter-based kernelized correlation filter (K2CF) algorithm. The adaptive Kalman filter can predict the state of linearly and nonlinearly fast-moving targets. The kernelized correlation filter algorithm then accurately detects the positions of the moving targets and uses the detection results to modify the moving states of the targets. This paper verifies the performance of the algorithm on a monocular vision measurement platform and using a pose measurement method. The K2CF algorithm was embedded in the micro-robot’s attitude measurement system, and the tracking performances of three different trackers were compared under different motion conditions. Our tracker improved the positioning accuracy and maintained real-time operation. In a comparison study of K2CF and many other algorithms on Object Tracking Benchmark-50 and Object Tracking Benchmark-100 video sequences, the K2CF algorithm achieved the highest accuracy. In the 400 mm × 300 mm field of view, when the target radius is about 3 mm and the inter-frame acceleration displacement does not exceed 5.6 mm, the root-mean-square error of position and attitude angle can satisfy the precision requirements of the system.

Sign in / Sign up

Export Citation Format

Share Document