Abstract
The detection and tracking of human moving objects is an important part of visual analysis of human movement and one of the important fields of computer vision. When using the existing methods to detect and track human track and field targets, there are some problems such as low detection accuracy, large target positioning error and low detection and tracking efficiency. A method of target detection and tracking in human track and field based on deep learning is proposed. The background subtraction method based on adaptive hybrid gaussian background model is used to detect the target. The read video image is denoised and smoothed. The holes in the foreground region are removed by morphological filtering, and the connected region of binary image is analyzed. Get the number and area of the connected areas. Human body area ratio and length-width ratio are used to classify and identify human body so as to complete the detection of human track and field sports target. Based on the structure of deep learning, combining the detection results of deep learning and LK tracking, PN learning was used to modify the parameters of the superposition automatic coding machine, avoiding the detection errors in deep learning and realizing the tracking of human track and field targets. Experimental results show that this method has higher detection accuracy, higher target positioning accuracy and higher detection and tracking efficiency.
Similar content being viewed by others
References
Chen TS, Chen JJ, Wu CH (2016) Distributed object tracking using moving trajectories in wireless sensor networks. Wirel Netw 22:1–23
Choi B, Park S, Kim E (2016) A newborn track detection and state estimation algorithm using Bernoulli random finite sets. IEEE Trans Signal Process 64:2660–2674
Geng PL, Song JC, Liu XF et al (2017) A method of mine moving objects detecting and tracking based on RGB color space. Journal of Taiyuan University of Technology 48:963–968
Gu SH, Lu B, Ma ZH (2017) Target detection and tracking algorithm based on contour and ASIFT feature. Computer Simulation 34:266–271
Gunes A, Guldogan MB (2018) Joint underwater target detection and tracking with the Bernoulli filter using an acoustic vector sensor. Digital Signal Processing 48:246–258
Kwon J, Kim K, Cho K (2017) Multi-target tracking by enhancing the kernelised correlation filter-based tracker. Electron Lett 53:1358–1360
Li L, Fan T, Zhen S et al (2016) Object tracking with double-dictionary appearance model. Opt Eng 55:103–106
Li F, Zhang R, Feng Y (2017) Fast pedestrian detection and dynamic tracking for intelligent vehicles within V2V cooperative environment. IET Image Process 11:833–840
Liu X, Chen SW, Aditya S et al. 2018. Robust Fruit Counting: Combining Deep Learning, Tracking, and Structure from Motion 26:156-158
Ma Y, Chang Q, Hu MF (2017) Research on infrared human detection from complex backgrounds. Infrared Technology 39:1038–1044
Marvasti FS, Mosavi MR, Nasiri M (2018) Flying small target detection in IR images based on adaptive toggle operator. IET Comput Vis 12:527–534
Park JD, Doherty JF (2017) Track detection of low observable targets using a motion model. IEEE Access 3:1408–1415
Shi J, Zhang SL (2017) Across first-order peak target detection based on RSMH for HF radar. Journal of China Academy of Electronics and Information Technology 12:25–30
Suhr JK, Jung HG (2016) Automatic parking space detection and tracking for underground and indoor environments. IEEE Trans Ind Electron 63:5687–5698
Thomas J, Loianno G (2017) Autonomous flight for detection, localization, and tracking of moving targets with a small quadrotor. IEEE Robotics & Automation Letters 99:1–1
Xia K-j, Yin H-s, Zhang Y-d (2019) Deep semantic segmentation of kidney and space-occupying lesion area based on SCNN and ResNet models combined with SIFT-flow algorithm. J Med Syst 43:2
Xue YD (2016) Application of binocular vision technology in human-computer interaction sports. Automation & Instrumentation 26:189–190
Xue H, Yao L, Deng C et al (2016) Tracking people in RGBD videos using deep learning and motion clues. Neurocomputing 204:70–76
Yi M, Chu Y (2016) Small moving target detection and tracking based on multi-view aerial video registration. Computer Engineering and Applications 52:27–31
Yuan Y, Lu Y, Qi W (2017). Tracking as a Whole: Multi-Target Tracking by Modeling Group Behavior With Sequential Detection. IEEE Transactions on Intelligent Transportation Systems PP:1-11
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Zhang, Y., Zhang, M., Cui, Y. et al. Detection and tracking of human track and field motion targets based on deep learning. Multimed Tools Appl 79, 9543–9563 (2020). https://doi.org/10.1007/s11042-019-08035-9
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-019-08035-9