Object Tracking using Modified Mean Shift Algorithm in a Robust Manner

Author(s):

Sadaf Khan , G H RAET NAGPUR; Dr. Sanjay Haridas, GHRAET, NAGPUR , MAHARASHTRA; Sanjay. B. Tembhurne, GHRAET, NAGPUR , MAHARASHTRA

Keywords:

Mean Shift, Local Binary Pattern, Tracking, Correlation, Color Histogram

Abstract:

Object tracking in real time videos is a critical task .So far many algorithms have been proposed to overcome the difficulties which arises from noise, occlusions and changes in the foreground object. But still there is a need for an efficient algorithm that can effectively track the object due to its varying position under various ambient conditions. Among various tracking methods, the mean shift tracking algorithm is a most widely used one due to its efficiency and simplicity. The experimental results show that the method which has been proposed can effectively track an object under the condition of varying illumination and shape deformation. Here the Correlation factor is also used to check the effective working of Mean shift algorithm and to make comparison between RGB and LBP technique.

Other Details:

Manuscript Id	:	IJSTEV2I1019
Published in	:	Volume : 2, Issue : 1
Publication Date	:	01/08/2015
Page(s)	:	17-24

Download Article