Title :
A region based CAMShift tracking with a moving camera
Author :
Kassir, Mohamad Mahdi ; Palhang, Maziar
Author_Institution :
Electr. & Comput. Eng. Dept., Artificial Intell. Lab., Iran
Abstract :
CAMShift algorithm is a fast and simple color based tracking algorithm but it is not robust to camera motion because it faces a problem when the tracked object moves across regions of background with similar colors with tracked object. Some algorithms have been proposed to reduce this problem. All of these proposed methods assign a probability to pixels belonging to object. However, in this paper we assign a probability to regions instead of pixels. Assigning probability to regions leads to a more accurate algorithm than the method of assigning probability to pixels. In this paper we proposed a region based approach to CAMShift (RCAMShift). We first demonstrate the problem of CAMShift. Then we show how our approach can improve CAMShift algorithm especially in a situation where there are similar colors in background.
Keywords :
image colour analysis; object tracking; RCAMShift algorithm; color based tracking algorithm; object tracking; region based CAMShift tracking; Algorithm design and analysis; Cameras; Color; Histograms; Image color analysis; Search problems; Tracking; ABCShift; CAMShift; Object tracking; RCAMShift; Robot vision;
Conference_Titel :
Robotics and Mechatronics (ICRoM), 2014 Second RSI/ISM International Conference on
Conference_Location :
Tehran
DOI :
10.1109/ICRoM.2014.6990943