Applying mean shift, motion information and Kalman filtering approaches to object tracking

Mazinan, A. H.; Amir-Latifi, A.

doi:10.1016/j.isatra.2012.02.002

Cited by 56 publications

(18 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this model, hue and shape are used as complementary features. One of the main parts of particle filter is dissemination algorithm of conditional probability density function [21][22][23][24][25][26][27][28][29][30][31][32][33]. An improved factored sampling algorithm is also realized to resample and weight particles.…”

Section: A Review Of Most Important Applied Methodsmentioning

confidence: 99%

A powerful approach to real-time mobile objects tracking in crowded environments utilizing 3D video analysis of fixed camera, particle filter and neural network

Mazinan

2014

Optik

View full text Add to dashboard Cite

Section: A Review Of Most Important Applied Methodsmentioning

confidence: 99%

A powerful approach to real-time mobile objects tracking in crowded environments utilizing 3D video analysis of fixed camera, particle filter and neural network

Mazinan

2014

Optik

View full text Add to dashboard Cite

“…The MIL, WMIL, and significance-MIL algorithms update their parameters with constant learning rate [5].

\begin{matrix} T17 \\ μ_{k}^{1} = λ μ_{k}^{1} + (1 - λ) μ^{1}, \\ σ_{k}^{1} = \sqrt{λ {(|, σ_{k}^{normal1})}^{normal2} + ((), normal1 - λ) {(|, σ^{normal1})}^{normal2} + λ ((), normal1 - λ) {(|, μ_{k}^{normal1} - μ^{normal1})}^{normal2}}, \\ μ_{k}^{0} = λ μ_{k}^{0} + (1 - λ) μ^{0}, \\ σ_{k}^{0} = \sqrt{λ {(|, σ_{k}^{normal0})}^{normal2} + ((), normal1 - λ) {(|, σ^{normal0})}^{normal2} + λ ((), normal1 - λ) {(|, μ_{k}^{normal0} - μ^{normal0})}^{normal2}}, \end{matrix}

where 0 < λ < 1 is the learning rate.…”

Section: Patch Based Mil Algorithmmentioning

confidence: 99%

Patch Based Multiple Instance Learning Algorithm for Object Tracking

Wang¹,

Wang²,

Zhang

2017

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

To deal with the problems of illumination changes or pose variations and serious partial occlusion, patch based multiple instance learning (P-MIL) algorithm is proposed. The algorithm divides an object into many blocks. Then, the online MIL algorithm is applied on each block for obtaining strong classifier. The algorithm takes account of both the average classification score and classification scores of all the blocks for detecting the object. In particular, compared with the whole object based MIL algorithm, the P-MIL algorithm detects the object according to the unoccluded patches when partial occlusion occurs. After detecting the object, the learning rates for updating weak classifiers' parameters are adaptively tuned. The classifier updating strategy avoids overupdating and underupdating the parameters. Finally, the proposed method is compared with other state-of-the-art algorithms on several classical videos. The experiment results illustrate that the proposed method performs well especially in case of illumination changes or pose variations and partial occlusion. Moreover, the algorithm realizes real-time object tracking.

show abstract

“…The mean-shift algorithm is a fast and accurate object tracking algorithm. It is a kind of non parametric algorithm, which was originally used in the pattern recognition as an cluster analysis, and is widely used in the field of object tracking [6][7][8][9][10]. Mean-shift algorithm is generally refers to an iterative steps, which calculates the mean shift, moving the point to the next mean shift, then taking this as the new starting point, and continue to move until it meets certain conditions.…”

Section: Tracking Algorithmmentioning

confidence: 99%

Real-time system design for object tracking

Liu

Li²,

Sang³

et al. 2016

Proceedings of the 2016 3rd International Conference on Mechatronics and Information Technology

View full text Add to dashboard Cite

This paper describes the realization of a system capable of tracking objects. The whole system design includes lower control system design and upper control system design. The rotation of the camera was controlled by an MCU in the lower control system and the tracking algorithm mean-shift was performed in a computer in the upper control system. Experiment results showed that the system could run well as a real time tracking system, and could track the object selected well.

show abstract

Applying mean shift, motion information and Kalman filtering approaches to object tracking

Cited by 56 publications

References 13 publications

A powerful approach to real-time mobile objects tracking in crowded environments utilizing 3D video analysis of fixed camera, particle filter and neural network

A powerful approach to real-time mobile objects tracking in crowded environments utilizing 3D video analysis of fixed camera, particle filter and neural network

Patch Based Multiple Instance Learning Algorithm for Object Tracking

Real-time system design for object tracking

Contact Info

Product

Resources

About