Digital Signal Processing Reference
In-Depth Information
Gaussian Particle Filter Based Algorithm for Tracking
of a Dim Moving Point Target in IR Image Sequences
Dilmurat Tursun and Askar Hamdulla
Institute of Information Science and Engineering, Xinjiang University, Urumqi, 830046, China
askar@xju.edu.cn
Abstract.
In order to solve the problems of deviations with large error occurred
at the sharp corners of motion trajectories and poor stability of tracking algo-
rithm presented in previous paper, and further make improvements to the track-
ing precision, in this paper the rather appropriate state model is established first
of all, and then the effective observations are collected by using spatial-
temporal detection and fusion in which the brightness information of the target
included in the decision criteria for the first time. Under the different circums-
tances of Gaussian noise and non-Gaussian noise, the experimental results show
that take the Gaussian particle filter as the tracking algorithm which has no use
of re-sampling could have effectively improved the precision of algorithm for
tracking dim moving point target in IR image sequences, and has a good real-
time performances and good stability.
Keywords:
target detection and tracking, IR image sequences, dim moving
point target, Gaussian particle filter.
1
Introduction
Dim moving target tracking refers to the process of analyzing the motion states of the
target formed by Infrared thermal imaging. Steady, high precision algorithm for track-
ing of IR target has a very important application prospect in military research field
such as guidance, reconnaissance etc. The essence of the target tracking procedure is
determination process of the target's positions in IR image sequences. In the case of
low SNR Infrared image sequences, the infrared dim point target contains less infor-
mation, no texture, no shape, no size features are available, resulting in dim target
detection and tracking was not accurate enough indeed. In the paper [1, 2, 3], accom-
plish the target tracking task based on probabilistic data association particle filter, but
found that the tracking trajectory in the corner will appear deviation, tracking errors
are large, algorithm's stability is not good enough.
On the analysis of the existing literatures, in order to solve the problem of target
tracking error, first of all, this paper established a more appropriate state model for
target movement. Then, after going through the process of space-time domain deci-
sion fusion according to the characteristics of target, and combined with the target's
intensity information get the final effective measurements, then apply the Gaussian
particle filter algorithm for real-time target tracking.
