Title :
A MeanShift-Particle Fusion tracking algorithm based on SIFT
Author :
Kai Zhou ; Rui-Xia Fan ; Wei-Xing Li
Author_Institution :
Sch. of Autom., Beijing Inst. of Technol., Beijing, China
Abstract :
In order to address vulnerability of color histogram to similar background and similar objects, a MeanShift-Particle amalgamative algorithm based on SIFT is set up, which takes advantage of Spatio Histogram and SIFT descriptors. According to the definition of SIFT match rate parameters, traditional particle filter algorithm, SIFT-particle filter algorithm and MeanShift algorithm are amalgamated effectively. The results of experiments showed the proposed amalgamative algorithm performs well when the object is cluttered.
Keywords :
image colour analysis; image fusion; image matching; object detection; particle filtering (numerical methods); statistical analysis; transforms; SIFT match rate parameters; SIFT-particle filter algorithm; amalgamative algorithm; color histogram; mean shift-particle fusion tracking algorithm; scale invariant feature transform; spatiohistogram; Computer vision; Electronic mail; Filtering theory; Head; Histograms; Kalman filters; Particle filters; Meanshift; Particle Filter; SIFT; Spatio Histogram;
Conference_Titel :
Control Conference (CCC), 2010 29th Chinese
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-6263-6
