Infrared Dim-Small Target Tracking Based on Guided Image Filtering and Kernelized Correlation Filtering

Dong Zhao; Huixin Zhou; Hanlin Qin; Kun Qian; Shenghui Rong; Kuanhong Cheng; Shangzhen Song

doi:10.3788/AOS201838.0204004

Journals >Acta Optica Sinica >Volume 38 >Issue 2 >Page 0204004 > Article

Acta Optica Sinica
Vol. 38, Issue 2, 0204004 (2018)

Infrared Dim-Small Target Tracking Based on Guided Image Filtering and Kernelized Correlation Filtering

Dong Zhao, Huixin Zhou^*, Hanlin Qin, Kun Qian, Shenghui Rong, Kuanhong Cheng, and Shangzhen Song

Author Affiliations

School of Physics and Optoelectronic Engineering, Xidian University, Xi'an, Shaanxi 710071, China

show less

DOI: 10.3788/AOS201838.0204004 Cite this Article Set citation alerts

Dong Zhao, Huixin Zhou, Hanlin Qin, Kun Qian, Shenghui Rong, Kuanhong Cheng, Shangzhen Song. Infrared Dim-Small Target Tracking Based on Guided Image Filtering and Kernelized Correlation Filtering[J]. Acta Optica Sinica, 2018, 38(2): 0204004 Copy Citation Text

show less

Fig. 1. Schematic of GIF processing. (a) Original image; (b) result by GIF processing; (c) residual image

Download full size

Fig. 2. Three-dimensional views correspongding to Figs. 1(a)~(c)

Download full size

Fig. 3. Tracking results under different backgrounds. (a) Bright cloud; (b) thick cloud 1; (c) thick cloud 2; (d) plane 1; (e) plane 2; (f) plane 3

Download full size

Fig. 4. Center location errors under different backgrounds. (a) Bright cloud; (b) thick cloud 1; (c) thick cloud 2; (d) plane 1; (e) plane 2; (f) plane 3

Download full size

Item	Bright colud	Thick cloud 1	Thick cloud 2	Plane 1	Plane 2	Plane 3
Image size /(pixel×pixel)	320×320	560×400	400×400	256×256	320×240	570×320
Sequence length /frame	300	340	300	120	500	375
Object size /(pixel×pixel)	3×3	3×3	3×3	2×2-6×6	3×3-7×7	3×3-7×7
Noise scale	Large	Very large	Very large	Very large	Large	Little
Speed /(pixel·frame^-1)	0.5	1	1	0.4-0.8	0.4-0.8	0.4-0.8
Motion trail	Curve	Straight	Straight	-	-	-

Table 1. Statistics of experimental image sequence candidates

Item	MPF	KCF	MS-PF	WCF	GIF-STC	GIF-KCF
R_S for bright cloud	0.16	0.62	0.59	0.63	1	1
R_S for thick cloud 1	0.15	0.13	0.24	0.53	1	1
R_S for thick cloud 2	0.25	0.01	0	0.01	1	1
R_S for plane 1	0.21	0.52	0.53	0.60	0.53	1
R_S for plane 2	0	0.53	0.83	0.52	0.85	0.99
R_S for plane 3	0	0.54	0.63	0.53	1	1
$\begin{matrix} {\bar{R}}_{S} \end{matrix}$	0.13	0.39	0.47	0.47	0.90	0.99
$\begin{matrix} {\bar{f}}_{frame}_{} \end{matrix}$ /(frame·s^-1)	6	72	25	33	52	71

Table 2. Performance comparison of tracking algorithms

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information