An enhanced Probability based object tracking technique for Multiple Object Tracking

  Er Harry,  Er Deepinder Kaur

Object tracking is an important step used in many computer vision applications to track targets like vehicles, human, birds etc. in an environment. Object tracking is the technique which is used to track object from the image or from the video. The video consists of multiple frames and in each frame we find the location of that object. To find the location of the object technique of probability has been applied and this technique the location of the object is predicted based on the probability density which works well on single and multiple objects. In this paper, improvement has been proposed in probability based technique to track multiple objects from the video which uses a combination of particle filter , kalman filter and frame differencing method considering the issues like degeneracy, non-linearity, partial occlusion. In this proposed technique the object showing the maximum variation will be tracked automatically based on the movement of the object without human intervention. The proposed technique is implemented in MATLAB. We analyzed the difficulties in video object tracking and tried to overcome these with more accuracy and less computation time. The enhanced method has shown better results based on the selected object tracking parameters

Keywords- : Object Tracking, Probabilistic approach, partial occlusion, multiple targets, particle filter, kalman filter, sequential importance resampling

Unique Identification Number - IJEDR1504095
Page Number(s) - 570-576
Pubished in - Volume 3 | Issue 4 | December 2015
DOI (Digital Object Identifier) -   
Publisher - IJEDR (ISSN - 2321-9939)

  Er Harry,  Er Deepinder Kaur,   "An enhanced Probability based object tracking technique for Multiple Object Tracking", International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.3, Issue 4, pp.570-576, December 2015, Available at :http://www.ijedr.org/papers/IJEDR1504095.pdf