Action Prediction and Scale Adaptive Target Tracking Algorithm

Xuemeng Tang; Zhiguo Chen; Yi Fu

doi:10.3788/LOP57.081014

Journals >Laser & Optoelectronics Progress >Volume 57 >Issue 8 >Page 081014 > Article

Laser & Optoelectronics Progress
Vol. 57, Issue 8, 081014 (2020)

Action Prediction and Scale Adaptive Target Tracking Algorithm

Xuemeng Tang, Zhiguo Chen^*, and Yi Fu

Author Affiliations

[in Chinese]

show less

DOI: 10.3788/LOP57.081014 Cite this Article Set citation alerts

Xuemeng Tang, Zhiguo Chen, Yi Fu. Action Prediction and Scale Adaptive Target Tracking Algorithm[J]. Laser & Optoelectronics Progress, 2020, 57(8): 081014 Copy Citation Text

show less

Fig. 1. Predicting car movement

Download full size

Fig. 2. Prediction results. (a) x coordinate prediction; (b) y coordinate prediction

Download full size

Fig. 3. Block detection results

Download full size

Fig. 4. Flow chart of algorithm

Download full size

Fig. 5. Flow chart of improved algorithm

Download full size

Fig. 6. Partial experimental results. (a) BlurCar; (b) Box; (c) Boy; (d) Car4; (e) Singer1; (f) Surfer; (g) Walking2; (h) Walking

Download full size

Fig. 7. Precision curves of partial experimental result. (a) Bolt; (b) Box; (c) Car2; (d) Dancer; (e) Human6; (f) Soccer

Download full size

Fig. 8. Error curves of partial experimental result. (a) Bolt; (b) Box; (c) Car2; (d) Dancer; (e) Human6; (f) Soccer

Download full size

Fig. 9. Test results. (a) Average CLE (front 250 frame); (b) DP graph

Download full size

Video	Numberof frames	Main factorof influence
BlurCar2	576	SV,MB,FM
Bolt	350	SV,DEF,BC,OCC,IPR,OPR
Bolt2	293	SV,DEF,BC
Box	400	SV,IV,OCC,MB,IPR,OPR,OV,BC,LR
Boy	602	SV,MB,FM,IPR,OPR
Car2	913	SV,MB,FM,BC,IV
Car4	659	SV,IV,BC
Dancer	225	SV,DEF,IPR,OPR
Doll	1000	IV,SV,OCC,IPR,OPR
Human6	792	SV,OCC,DEF,FM,OPR,OV
Human8	128	IV,SV,DEF
Singer1	319	IV,SV,OCC,OPR
Soccer	380	IV,SV,OCC,MB,FM,IPR,OPR,BC
Surfer	370	SV,FM,IPR,OPR,LR
Walking	412	SV,OCC,DEF
Walking2	500	SV,OCC,LR
Woman	597	SV,OCC, IV,DEF,MB,FM,OPR

Table 1. Test video in the experiment

Algorithm	DSST	ECO	KCF	LADCF	LDES	SAMF	FSKCF
Bolt	100.00	100.00	98.90	100.00	100.00	100.00	100.00
Bolt2	2.10	83.30	1.70	39.90	01.70	1.70	100.00
Boy	100.00	100.00	100.00	100.00	100.00	100.00	100.00
Car4	100.00	100.00	95.00	100.00	99.80	100.00	100.00
Human6	44.80	98.90	29.00	97.90	98.90	92.30	99.70
Surfer	70.00	100.00	90.80	100.00	100.00	100.00	100.00
Average	69.48	97.03	69.23	89.63	83.40	82.33	99.95

Table 2. Tracking precision of the algorithm on fast motion video%

Algorithm	DSST	ECO	KCF	LADCF	LDES	SAMF	FSKCF
BlurCar2	100.0	100.0	97.9	100.0	100.0	100.0	100.0
Bolt	100.0	100.0	98.9	100.0	100.0	100.0	100.0
Bolt2	2.1	83.3	1.7	39.9	1.7	1.7	100.0
Box	100.0	99.8	100.0	100.0	98.8	100.0	100.0
Boy	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Car2	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Car4	100.0	100.0	95.0	100.0	99.8	100.0	100.0
Dancer	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Doll	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Human6	44.8	98.9	29.0	97.9	98.9	92.3	99.7
Human8	100.0	100.0	100.0	100.0	32.8	100.0	100.0
Singer1	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Soccer	70.5	24.0	81.3	21.1	24.7	20.8	97.4
Surfer	70.0	100.0	90.8	100.0	100.0	100.0	100.0
Walking	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Walking2	100.0	100.0	44.0	100.0	100.0	100.0	100.0
Woman	93.8	100.0	93.8	100.0	98.5	93.8	99.2

Table 3. Tracking precision of the algorithm on the video%

Algorithm	DSST	ECO	KCF	LADCF	LDES	SAMF	FSKCF
BlurCar2	1.00	1.00	0.98	1.00	1.00	1.00	1.00
Bolt	1.00	1.00	0.99	1.00	1.00	1.00	0.99
Bolt2	0.01	0.78	0.01	0.30	0.01	0.01	0.23
Box	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Boy	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Car2	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Car4	1.00	1.00	0.51	1.00	1.00	1.00	1.00
Dancer	1.00	1.00	1.00	1.00	1.00	1.00	0.96
Doll	1.00	1.00	0.90	1.00	1.00	1.00	0.86
Human6	0.46	0.99	0.24	0.99	1.00	0.47	0.28
Human8	1.00	1.00	0.73	1.00	0.29	0.78	1.00
Singer1	1.00	1.00	0.44	1.00	1.00	0.75	1.00
Soccer	0.57	0.23	0.44	0.24	0.23	0.21	0.58
Surfer	0.36	0.97	0.74	1.00	1.00	1.00	0.74
Walking	1.00	1.00	0.73	1.00	1.00	1.00	1.00
Walking2	1.00	1.00	0.43	1.00	1.00	0.99	1.00
Woman	0.94	0.93	0.94	0.94	0.96	0.94	0.98

Table 4. Overlap rate of the algorithm tracking on the video

Parameter	DSST	ECO	KCF	LADCF	LDES	SAMF	FSKCF
DP/%	92.4	95.2	89.4	94.9	90.8	94.2	99.8
CLE	21.7	11.5	33.0	12.5	36.4	18.5	2.7
OP	0.87	0.93	0.74	0.91	0.84	0.85	0.88
V /(frame/s)	32.7	44.3	224.6	12.8	11.0	15.0	66.7

Table 5. Mean values of indicators on 17 videos by different algorithms

Xuemeng Tang, Zhiguo Chen, Yi Fu. Action Prediction and Scale Adaptive Target Tracking Algorithm[J]. Laser & Optoelectronics Progress, 2020, 57(8): 081014

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information

微信扫一扫：分享

微信扫一扫：分享