To solve the problem that SiamFC (fully-convolutional siamese network) has fast-tracking speed but is easily disturbed by motion blur, we proposed an algorithm that can effectively track objects in motion blur scene by introducing IQA (image quality assessment) module and deblurring module to improve SiamFC’s performance. On the one hand, we use the IQA module to evaluate the search area of SiamFC. Only when the results show that the image area is fuzzy, it can be processed by the deblurring module, which is to prevent introducing noise into the clear image. At the same time, the image quality assessment module and the network structure of SiamFC are integrated to minimize the amount of calculation required for the new module. On the other hand, the deblurring module can effectively solve the problem that the motion blur in the original image interferes with the tracking. The experimental results on the public data sets OTB2015 and UAV20L show that compared with the benchmark algorithm SiamFC, the precision rate of this algorithm on OTB2015 has increased by 3.1%, and the success rate has increased by 2.6%; the precision rate on UAV20L has increased by 4.7%, the success rate increased by 5.4%. The analysis of video sequences with different attributes in the OTB2015 and UAV20L datasets shows that the proposed algorithm can effectively reduce the interference of dynamic blur on the tracking algorithm.
WUY, LIM J, YANGM H. Object tracking benchmark[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2015, 37(9):1834-1848. DOI:10.1109/TPAMI.2014.2388226 .
[2]
SONGY B, MAC, WUX H, et al. VITAL: Visual tracking via adversarial learning [C]//2018 Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE Press,2018: 8990-8999. DOI: 10.1109/CVPR.2018.00937 .
[3]
DINGJ W, HUANGK Q, TANT N. Tracking blurred object with data-driven tracker [C]// 2012 IEEE Ninth International Conference on Advanced Video and Signal⁃Based Surveillance. New York: IEEE Press,2012: 331-336. DOI: 10.1109/AVSS.2012.78 .
[4]
YIW, LINGH B, YUJ Y, et al. Blurred target tracking by blur-driven tracker[C]// 2011 International Conference on Computer Vision. New York: IEEE Press, 2011: 1100-1107. DOI: 10.1109/ICCV.2011.6126357 .
[5]
YUANL, SUNJ, QUANL, et al. Image deblurring with blurred/noisy image pairs[C]//ACM SIGGRAPH 2007 Papers on-SIGGRAPH’07. New York: ACM Press, 2007: 1-11. DOI:10.1145/1275808.1276379 .
[6]
JINH, FAVAROP, CIPOLLAR. Visual tracking in the presence of motion blur[C]//2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2005: 18-25. DOI:10.1109/CVPR.2005.372 .
[7]
DAIS Y, YANGM, WUY, et al. Tracking motion-blurred targets in video[C]// 2006 International Conference on Image Processing. New York: IEEE Press, 2006: 2389-2392. DOI: 10.1109/ICIP.2006.312943 .
[8]
DINGJ W, HUANGY Z, LIUW, et al. Severely blurred object tracking by learning deep image representations[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2016, 26(2): 319-331. DOI:10.1109/TCSVT.2015.2406231 .
[9]
KUPYNO, BUDZANV, MYKHAILYCHM, et al. DeblurGAN: Blind motion deblurring using conditional adversarial networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern RecognitionSalt Lake City. New York: IEEE Press, 2018 : 8183-8192. DOI: 10.1109/CVPR.2018.00854 .
[10]
KUPYNO, MARTYNIUKT, WUJ R, et al. DeblurGAN-v2: Deblurring (orders-of-magnitude) faster and better [C]//2019 Proceedings of the IEEE/CVF International Conference on Computer Vision. New York: IEEE Press, 2019: 8878-8887. DOI:10.1109/ICCV.2019.00897 .
[11]
LINT Y, DOLLÁRP, GIRSHICKR, et al. Feature pyramid networks for object detection[C]//2017 IEEE Proceedings of the Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2017: 2117-2125. DOI: 10.1109/cvpr.2017.106 .
[12]
TAOX, GAOH Y, SHENX Y, et al. Scale-recurrent network for deep image deblurring [C]//2018 Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2018: 8174-8182. DOI:10.1109/CVPR.2018.00853 .
[13]
WANGX T, CHANK C K, YUK, et al. EDVR: Video restoration with enhanced deformable convolutional networks [C]//2019 Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. New York: IEEE Press, 2019: 1954-1963. DOI:10.1109/CVPRW.2019.00247 .
[14]
BERTINETTOL, VALMADREJ, HENRIQUESJ F, et al. Fully-convolutional siamese networks for object tracking [C]//2016 European Conference on Computer Vision. Berlin: Springer, 2016: 850-865.DOI: 10.1007/978-3-319-48881-3_56 .
[15]
LIUX L, DE WEIJER JVAN, BAGDANOVA D. RankIQA: Learning from rankings for no-reference image quality assessment [C]//2017 IEEE International Conference on Computer Vision. New York: IEEE Press, 2017: 1040-1049. DOI: 10.1109/ICCV.2017.118 .
MUELLERM, SMITHN, GHANEMB. A benchmark and simulator for uav tracking [C]//2016 European Conference on Computer Vision. Berlin: Springer, 2016: 445-461. DOI:10.1007/978-3-319-46448-0_27 .
[18]
KRIZHEVSKYA, SUTSKEVERI, HINTONG E. ImageNet classification with deep convolutional neural networks [J]. Advances in neural information processing systems, 2012, 25(1): 1097-1105. DOI: 10.1145/3065386 .
[19]
SIMONYANK, ZISSERMANA. Very deep convolutional networks for large-scale image recognition [C]//2014 International Conference on Learning Representations. New York: IEEE Press, 2014: 1-15. DOI:10.1109/ICLR.2014.7486599 .
[20]
SHIJ P, XUL, JIAJ Y. Discriminative blur detection features [C]//2014 IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2014: 2965-2972. DOI:10.1109/CVPR.2014.379 .
[21]
BROOKST, BARRONJ T. Learning to synthesize motion blur [C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE Press, 2019: 6840-6848. DOI:10.1109/CVPR.2019.00700 .
[22]
LIF F, FERGUSR, PERONAP. Learning generative visual models from few training examples: An incremental bayesian approach tested on 101 object categories[C]//2004 IEEE Conference on Computer Vision and Pattern Recognition Workshop. New York: IEEE Press, 2004: 178-178. DOI: 10.1109/CVPR.2004.383 .
[23]
DANELLJANM, HÄGERG, KHANF S, et al. Accurate scale estimation for robust visual tracking [C]// 2014 British Machine Vision Conference. Nottingham: BMVA Press, 2014: 1-11. DOI: 10.5244/C.28.65 .
[24]
BIBIA, MUELLERM, GHANEMB. Target response adaptation for correlation filter tracking [C]//2016 European Conference on Computer Vision. Berlin:Springer, 2016: 419-433. DOI:10.1007/978-3-319-46466-4_25 .
[25]
LIY, ZHUJ K. A scale adaptive kernel correlation filter tracker with feature integration[C]//2014 European Conference on Computer Vision. Berlin:Springer, 2014: 254-265. DOI: 10.1007/978-3-319-16181-5_18 .
[26]
HENRIQUESJ F, CASEIROR, MARTINSP, et al. High-speed tracking with kernelized correlation filters[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 37(3): 583-596. DOI: 10.1109/TPAMI.2014.2345390 .
[27]
DANELLJANM, HÄGERG, KHANF S, et al. Learning spatially regularized correlation filters for visual tracking [C]//2015 IEEE International Conference on Computer Vision. New York: IEEE Press, 2015: 4310-4318. DOI: 10.1109/ICCV.2015.490 .