Improved mean shift for multi-target tracking

  title={Improved mean shift for multi-target tracking},
  author={G. Phadke and R. Velmurugan},
  journal={2013 IEEE International Workshop on Performance Evaluation of Tracking and Surveillance (PETS)},
  • G. Phadke, R. Velmurugan
  • Published 2013
  • Geography
  • 2013 IEEE International Workshop on Performance Evaluation of Tracking and Surveillance (PETS)
Object tracking is critical to visual surveillance and activity analysis. The color based mean shift has been addressed as an effective and fast algorithm for tracking. But it fails in case of objects with low color intensity, clutter in background and total occlusion for several frames. We present a new scheme based on multiple feature integration for visual tracking. The proposed method integrates the color, texture and edge features of the target to construct the target model and a… Expand

Figures and Tables from this paper

Improved mean shift target tracking based on self-organizing maps
Thanks to its simplicity and real-time processing possibility, mean shift has been widely used for video tracking. However, it often fails when the background is similar to the intended object orExpand
A robust object tracking algorithm based on SURF
  • Dan Zhou, D. Hu
  • Computer Science
  • 2013 International Conference on Wireless Communications and Signal Processing
  • 2013
A robust object tracking algorithm based on SURF is proposed which uses a two-stage matching method to improve the accuracy of SURF matching points and uses matching points between new template and candidate region to locate the initial position of object. Expand
People Detection and Tracking in Low Frame-rate Dynamic Scenes
Results show that tracking is very dependent on the frame rate while detection is much more robust in this scenario, and reducing the video frame rate can be an alternative to reduce the amount of generated data. Expand
Moving human full body and body parts detection, tracking, and applications on human activity estimation, walking pattern and face recognition
A super-resolution algorithm is applied to enhance the face resolution for improved human face recognition performance and a new algorithm to split a merged detection blob back to individual detections based on the detected head positions is developed. Expand
Non-overlapping horizon multi-target recognition based on multi-source information fusion
Aiming at the complexity and diversity of real-time target tracking in complex electromagnetic environment, a method of target recognition and tracking based on scale invariant feature is proposed byExpand


Robust object tracking with background-weighted local kernels
A weighted fragment based approach that tackles partial occlusion is proposed that is computationally simple enough to be executed in real-time and can be directly extended to a multiple object tracking system. Expand
Robust multiple target tracking under occlusion using fragmented mean shift and Kalman filter
  • G. Phadke
  • Computer Science
  • 2011 International Conference on Communications and Signal Processing
  • 2011
This paper investigates how to improve the robustness of visual tracking method for multiple target tracking with occlusion handling and proposes weighted fragment based mean shift with Kalman filter with the consideration of color features of the target. Expand
Improving of Mean Shift Tracking Algorithm Using Adaptive Candidate Model
This paper proposes Improving of Mean Shift Tracking Algorithm Using Adaptive Candidate (MST_AC) Model, which uses the background positions on the previous frame and the current frame to compute the new candidate model. Expand
Mean shift-based object tracking with multiple features
It is shown that tracking with multiple weighted features provides more reliable performance than single features tracking. Expand
An improved Mean-shift tracking algorithm based on color and texture feature
Aimed at the defect of classic Mean-shift tracking algorithm which is vulnerable to similar background inference for using single color feature,an improved color and texture features combinedExpand
Real-time tracking of non-rigid objects using mean shift
The theoretical analysis of the approach shows that it relates to the Bayesian framework while providing a practical, fast and efficient solution for real time tracking of non-rigid objects seen from a moving camera. Expand
An approach based on mean shift and KALMAN filter for target tracking under occlusion
This paper combines the mean shift algorithm with the Kalman filer for target tracking, and shows that the proposed algorithm is very effective to solve the occlusion problem. Expand
Integrating Color and Shape-Texture Features for Adaptive Real-Time Object Tracking
The standard mean-shift tracking algorithm is extended to an adaptive tracker by selecting reliable features from color and shape-texture cues according to their descriptive ability, which makes the tracker more robust. Expand
Robust and efficient fragments-based tracking using mean shift
Abstract Representing a target with multiple image fragments has been used for tracking in a video sequence and proved to be able to preserve the spatial information. However, this kind of targetExpand
Fragment Based Tracking for Scale and Orientation Adaptation
In this work, we propose a simple yet highly effective algorithm for tracking a target through significant scale and orientation change. We divide the target into a number of fragments and trackingExpand