Citation: Huo, Y.; Chen, Y.; Zhang,
H.; Zhang, H.; Wang, H. Dim and
Small Target Tracking Using an
Improved Particle Filter Based on
Adaptive Feature Fusion. Electronics
2022, 11, 2457. https://doi.org/
10.3390/electronics11152457
Academic Editor: Silvia Liberata Ullo
Received: 13 July 2022
Accepted: 5 August 2022
Published: 7 August 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Article
Dim and Small Target Tracking Using an Improved Particle
Filter Based on Adaptive Feature Fusion
Youhui Huo
1,2
, Yaohong Chen
1
, Hongbo Zhang
3
, Haifeng Zhang
1,
* and Hao Wang
1,
*
1
Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an 710119, China
2
University of Chinese Academy of Sciences, Beijing 100049, China
3
China Astronaut Research and Training Center, Beijing 100094, China
* Correspondence: zhanghf@opt.ac.cn (H.Z.); wanghao@opt.ac.cn (H.W.)
Abstract:
Particle filters have been widely used in dim and small target tracking, which plays
a significant role in navigation applications. However, their characteristics, such as difficulty of
expressing features for dim and small targets and lack of particle diversity caused by resampling,
lead to a considerable negative impact on tracking performance. In the present paper, we propose an
improved resampling particle filter algorithm based on adaptive multi-feature fusion to address the
drawbacks of particle filters for dim and small target tracking and improve the tracking performance.
We first establish an observation model based on the adaptive fusion of the features of the weighted
grayscale intensity, edge information, and wavelet transform. We then generate new particles based
on residual resampling by combining the target position in the previous frame and the particles in
the current frame with higher weights, with the tracking accuracy and particle diversity improving
simultaneously. The experimental results demonstrate that our proposed method achieves a high
tracking performance with a distance accuracy of 77.2% and a running speed of 106 fps, respectively,
meaning that it will have a promising prospect in dim and small target tracking applications.
Keywords: dim and small target; target tracking; feature fusion; particle filter; resampling method
1. Introduction
Video processing technologies such as target tracking [
1
,
2
], target detection [
3
,
4
], and
moving target segmentation [
5
,
6
] have made great progress, with related applications
continuing to emerge. Among them, target tracking has been widely used in intelligent
video surveillance, modern military, and intelligent visual navigation equipment. A set of
conventional methods have been proposed to meet the application requirements, including
particle filters (PF) [
7
], kernelized correlation filters (KCF) [
8
], efficient convolution opera-
tors (ECO) [
9
], and spatial-temporal regularized correlation filters (STRCF) [
10
]. Dim and
small target tracking is an important branch of target tracking, which plays a key role in
military, aviation, and aerospace fields such as image matching guidance and reconnais-
sance [
2
]. Dim and small target tracking possesses more stringent requirements for trackers
because the targets occupy only a small number of pixels in the image (the target size is
generally about 2
×
2), which means that there is a lack of feature information, such as
shape and texture, and a sensitivity to noise [11].
Recently, scholars have proposed many tracking algorithms to meet the requirements
of the dim and small target tracking, which can be divided into two main categories
including correlation filters-based methods [
12
–
14
] and PF-based methods [
15
–
19
]. KCF
has received the most attention in the methods based on correlation filters for dim and small
target tracking, which is achieved by establishing a discriminator based on the correlation
operator with a kernel function. Qian K. et al. proposed an anti-interference small target
tracking algorithm [
12
], which combines KCF with a detection model (KCFD), making
a robust tracking result on image sequences with complex backgrounds. Zhang L. et al.
Electronics 2022, 11, 2457. https://doi.org/10.3390/electronics11152457 https://www.mdpi.com/journal/electronics
