Abstract
In the process of image acquisition, the existing image-based real-time video acquisition system is susceptible to noise and distortion due to the influence of attitude, illumination and other conditions, which reduces the quality and stability of the acquired image, and thus makes it difficult to locate the image feature area. Therefore, the feature extraction technology of real-time video capture based on deep convolution neural network is proposed. Cut out high-quality images by locating reference points in feature connection areas, smooth each part of the image by using mean image filter, extract texture features by using convolution, transform, discrete cosine transform and statistical features, and replace random initialization weights with pre-trained models. In the process of model training and recognition, the methods of feature state division, image preprocessing and observation vector calculation are studied. The experimental results on ORL database verify the effectiveness of the image feature extraction method, which can meet the needs of current real-time video capture.
Similar content being viewed by others
References
Abderazek S, Nadia NT, Nadjib B et al (2019) A review on security challenges of wireless communications in disaster emergency response and crisis management situations. J Net Comput Appl
Bai JH (2017) Video Cloud. Protects urban safety. China Public Safety
Cai YF, Wang H, Sun XQ et al (2017) Scene adaptive pedestrian detection based on depth model. J Southeast Univ Nat Sci Ed
Fei L, Gao FJ, Wang YL (2015) Three-dimensional human body database query system. Auto Technol Appl
Gao QW, Fang HS (2017) HOG pedestrian detection algorithm based on multi-convolution feature fusion. Comput Sci
Guan W, Xue Y, An Z (2016) Application Progress and Prospect of Deep Learning in Video Target Tracking. J Auto
Hanxi L, Yi L, Fatih P (2016) Convolutional neural net bagging for online visual tracking. Comput Vis Image Underst
Hao TZ, Hong LH, Wen YZ (2018) Classifier selection and clustering with fuzzy assignment in ensemble model for credit scoring. Neurocomputing
Janghoon C, Junseok K, Kyoung ML (2018) Real-time visual tracking by deep reinforced decision making. Comput Vis Image Underst
Jiang T, Li ZH (2018) Review of local pattern mining in gene expression data. Comput Res Dev
Jingwei J, Ligang L, Zihao J et al (2018) A cellular automata model for high-density crowd evacuation using triangle grids. Phys A Stat Mech Appl
Lian W, He BW, Wang Y (2016) Research on video image synthesis application technology based on cloud computing technology. Police Technol
Liang XY (2016) ZynQ-based pedestrian detection system software and hardware collaborative implementation. Dalian Maritime University
Liu J (2014) Study on the Effect of Computer Vision and Machine Learning Technology in 3D Human Body Animation. Art Educ Res
Liu Q, Zhai JW, Zhang ZC et al (2018) Review of deep reinforcement learning. Chin J Comput
Liu WT, Lu XM (2018) Gait recognition based on layered fusion of LBP and HOG features. Comput Eng Appl
Luo HB, Xu L Yun, Hui B et al (2017) Research status and prospects of target tracking methods based on deep learning. Infrared Laser Eng
Qian J, Xin J, Shin JL et al (2018) A new similarity/distance measure between intuitionistic fuzzy sets based on the transformed isosceles triangles and its applications to pattern recognition. Expert Syst Appl
Sanaz B, Behzad G, Theodoros L et al (2019) A team-formation algorithm for faultline minimization. Expert Syst Appl
Shu CC, Li BF, Wei GS (2018) Exit selection and pedestrian movement in a room with two exits under fire emergency. Appl Math Comput
Sun R, Hou NQ, Chen J (2014) Fast pedestrian detection based on feature fusion and cross-core SVM. Optoelectronic Eng
Tingting S, Min M, Chenyu S et al (2018) Periodontal disease and susceptibility to breast cancer: A meta‐analysis of observational studies. J Clin Periodontol
Wang HF, Wang YB, Ke Y (2018) Research progress in three-dimensional anthropometric techniques. J Garment
Wang R, Zhang WL, Zhao MM et al (2016) Comparative analysis of three-dimensional anthropometric accuracy. Fash Des Eng
Wen XX (2018) Research and Design of Smart City Video Cloud Solution. Commun Power Technol
Xu Y, Xu XL, Li CN et al (2016) Pedestrian detection combined with SVM classifier and HOG feature extraction. Comput Eng
Yu JY, Ding PC, Wang C (2018) Summary of the application of convolutional neural networks in target detection. Comput Sci
Yu SP (2017) Analysis of the impact of 3D anthropometry application on the field of clothing. Western Leather
Yu ZQ, Qin J (2016) Research and design of video image information integrated management platform for incidents. Henan Sci
Yuki S, Kazuhiro I, Takayuki Y et al (2019) Data mining based on clustering and association rule analysis for knowledge discovery in multiobjective topology optimization. Expert Syst Appl
Zhang JH (2018) Research on pedestrian detection and tracking algorithm based on multi-scale method. Southwest Univ Sci Tech
Zhe X, Junping D, Dawei D et al (2019)Deep Low-Rank Subspace Ensemble for Multi-View Clustering. Inf Sci
Zeng XK, Lei Z, Kun L (2019) An improved social force model for pedestrian dynamics in shipwrecks. Appl Math Comput
Zhang JY, Wang HL, Guo Y et al (2018) Review of related research on deep learning. Comput Appl Res
Zhao WD, Zhao Y, Cheng SS (2018) Analysis of the overall architecture of public security video image information application system. Chin Sec Protect Cert
Acknowledgements
This work is partially supported by the Natural Science Fund of China (under Grant 61401356), the Special Fund for Technological Innovation Guidance of Shaanxi Province (under grant 2020CGXNG-015), Youth Innovation Team Building Scientific Research Project of Shaanxi Province (under Grant 21JP106), and the Science and Technology Project of Xi’an (under Grant 2019KJWL07, 2020kjrc0104).
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
Yang, S., Chong, X. Study on feature extraction technology of real-time video acquisition based on deep CNN. Multimed Tools Appl 80, 33937–33950 (2021). https://doi.org/10.1007/s11042-021-11417-7
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-021-11417-7