Proceedings Article10.1109/ICMA.2019.8816202
An Improved Struck Tracking Algorithm Based on Scale Adaptation and Selective Updating
Enzeng Dong,Mengtao Deng,Jigang Tong +2 more
- 01 Aug 2019
2
TL;DR: The experimental results demonstrate that the improved Struck algorithm outperforms the classical Struck in accuracy and robustness, and it is competitive with other state-of-the-art trackers on several challenging sequences.
read more
Abstract: In order to solve the problem of tracking failures caused by scale variation, occlusion and fast motion, a selective updating ASMS(scale-adaptive meanshift)-Struck algorithm which combines scale-adaptive meanshift and Perceptual hash method is proposed. By introduced scale-adaptive, the improved method can effectively track targets with scale variation and fast motion; then, the perceptual hash algorithm and Hamming distance are introduced to select a good update module in real time, which effectively avoids the error information introduced by the classifier due to scale variation, occlusion or illumination variation, thus improving the tracking robustness and accuracy. OTB-50 standard datasets is used to verify and evaluate our method. The experimental results demonstrate that the improved Struck algorithm outperforms the classical Struck in accuracy and robustness, and it is competitive with other state-of-the-art trackers on several challenging sequences.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
An Active Security System Based on AR Smart Classes and Face Recognition Technology
C. Charles Chen,Xiao-Yan Wang Charles Chen,Jian Chen Xiao-Yan Wang,Qin Liu Jian Chen,Sheng-Lung Peng Qin Liu +4 more
TL;DR: This research combines AR smart glasses, multi face recognition technology and control system technology to develop a face recognition active security platform which can effectively improve the arrest of criminals and establish a safer social environment.
5
Anti-Occlusion and Anti-deformation Hand Tracking Based on Kernel Correlation Filter
Xiaolu Guo,Li'nan Jiao,Youquan Liu,Xiao Hu +3 more
TL;DR: A novel hand tracking algorithm is proposed, combining Kernel Correlation Filter with Histograms of Oriented Gradients and Color Name features, to improve anti-occlusion and anti-deformation performance, achieving higher tracking accuracy and success rate compared to existing algorithms.
References
High-Speed Tracking with Kernelized Correlation Filters
TL;DR: A new kernelized correlation filter is derived, that unlike other kernel algorithms has the exact same complexity as its linear counterpart, which is called dual correlation filter (DCF), which outperform top-ranking trackers such as Struck or TLD on a 50 videos benchmark, despite being implemented in a few lines of code.
Online Object Tracking: A Benchmark
Yi Wu,Jongwoo Lim,Ming-Hsuan Yang +2 more
- 23 Jun 2013
TL;DR: Large scale experiments are carried out with various evaluation criteria to identify effective approaches for robust tracking and provide potential future research directions in this field.
Visual object tracking using adaptive correlation filters
David S. Bolme,J. Ross Beveridge,Bruce A. Draper,Yui Man Lui +3 more
- 13 Jun 2010
TL;DR: A new type of correlation filter is presented, a Minimum Output Sum of Squared Error (MOSSE) filter, which produces stable correlation filters when initialized using a single frame, which enables the tracker to pause and resume where it left off when the object reappears.
Object Tracking Benchmark
TL;DR: An extensive evaluation of the state-of-the-art online object-tracking algorithms with various evaluation criteria is carried out to identify effective approaches for robust tracking and provide potential future research directions in this field.
Tracking-Learning-Detection
TL;DR: A novel tracking framework (TLD) that explicitly decomposes the long-term tracking task into tracking, learning, and detection, and develops a novel learning method (P-N learning) which estimates the errors by a pair of “experts”: P-expert estimates missed detections, and N-ex Expert estimates false alarms.
3.7K