Fast Adaptive Tracking Based on Fusion Particle Filter Algorithm
Download as PDF
DOI: 10.23977/meimie.2019.43001
Author(s)
Jubao Qu, Hongtao Liang
Corresponding Author
Jubao Qu
ABSTRACT
To lock and track the target quickly, a novel fast target tracking algorithm is proposed, which combines adaptive scale Harris corner detection target, SIFT adaptive matching target and particle filter algorithm. This algorithm can extract highly discriminant features of the target adaptively, and make the feature pair rotate and scale zoom. The brightness change can keep invariance, and maintain a certain degree of stability for affine transformation, angle change and noise. It can also recognize targets well in the case of confusion and occlusion. The simulation results show that the algorithm can still accurately identify and track targets under multiple external disturbances, and has certain application value.
KEYWORDS
Particle filter, Adaptive tracking, Harris, SIFT