Abstract
In order to identify the falling behavior quickly and accurately from the surveillance video, an optimized model of alphapose is proposed here. First, the pedestrian target detection model and pose estimation model are accelerated by the lightweight YOLOv4-tiny model that we adopted. Then, the human pose joint point coordinate data obtained by the alphapose model is used to judge the occurrence of a falling action by the proposed judging algorithm. We use relationship between the head joint point line velocity and the crotch joint line velocity at the moment of a human fall and the change of angle between the perpendicular bisector and the x-axis of the image to judge the occurrence of the fall action. The algorithm proposed in this paper was compared with the main human posture-based fall detection algorithms for comparative analysis, with an image resolution of 320 × 240. The results show that the model proposed in this paper can timely and accurately detect the occurrence of pedestrian fall behavior.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Yin, J., Yang, Q.: Sensor-based abnormal human-activity detection. IEEE Trans. Knowl. Data Eng. 20(8), 1092–1099 (2008)
Khan, A.M., Lee, Y.K., Lee, S.Y., Kim, T.S.: A Triaxial accelerometer-based physical- activity recognition via augmented-signal features and a hierarchical recognizer. Inf. Technol. Biomed. IEEE Trans. 14(5), 1166–1172 (2010)
Sigg, S., Scholz, M., Shi, S., Ji, Y.: RF-sensing of activities from non-cooperative subjects in device-free recognition systems using ambient and local signals. IEEE Trans. Mob. Comput. 13(4), 907–920 (2014)
Wang, X., Gao, L., Wang, P., Sun, X., Liu, X.: Two-stream 3-d convnet fusion for action recognition in videos with arbitrary size and length. IEEE Trans. Multimed. 20(3), 634–644 (2017)
Arif, S., Wang, J., Ul Hassan, T., Fei, Z.: 3D-CNN-based fused feature maps with LSTM applied to action recognition. Future Int. 11(2), 42 (2019)
Kong, Y., Huang, J., Huang, S., Wei, Z., Wang, S.: Learing spatiotemporal representations for human fall detection in surveillance video. J. Vis. Commun. Image Represent. 59, 215–230 (2019)
Espinosa, R., Ponce, H., Gutiérrez, S., Martínez-Villaseñor, L., Brieva, J., Moya-Albor, E.: A vision-based approach for fall detection using multiple cameras and convolutional neural networks: a case study using the UP-Fall detection dataset. Computer. Biol. Med. 115, 103520 (2019)
Xu, T., Zhou, Y.: Elders’ fall detection based on biomechanical features using depth camera. Int. J. Wavelets, Multires. Inf. Process. 16(02), 1840005 (2018)
Kaihoko, Y., Tan, P.X., Kamioka, E.: Prevention of unintended appearance in photos based on human behavior analysis. Information 11(10), 468 (2020)
Thacker, C.B., Makwana, R.M.: Human behavior analysis through facial expression recognition in images using deep learning. Int. J. Innov. Technol. Exp. Eng. 9(2), 391–397 (2019)
Xu, Q., Zheng, W., Song, Y., Zhang, C., Yuan, X., Li, Y.: Scene image and human skeleton-based dual-stream human action recognition. Pattern Recogn. Lett. 148, 136–145 (2021)
Athavale, V.A., Gupta, S.C.: Human action recognition using CNN-SVM model. Adv. Sci. Technol. 6258, 282–290 (2021)
Bulbul, M.F., Ali, H.: Gradient local auto-correlation features for depth human action recognition. SN Appl. Sci. 3(5), 1–13 (2021)
Saifuddin, S.F.M., Shaahriar, K.A.: Silhouette pose feature-based human action classification using capsule network. J. Inf. Technol. Res. 14(12), 106–124 (2021)
Zhdanova, M.M., Balabaeva, O.S.: Improvement of the human action recognition algorithm by the pre-processing of input data. In: IOP Conference Series: Materials Science and Engineering vol. 1029, no. 1, pp. 12119–12119 (2021)
Suresh, K.B., Viswanadha, R.S.: Human action recognition using a novel deep learning approach. In: IOP Conference Series: Materials Science and Engineering. Vol. 1042, no. 1, p. 012031 (2021)
Hao, X., Li, J., Jiang, T., Yu, M.: Hypergraph neural network for skeleton-based action recognition. IEEE Trans. Image Process. 30, 2263–2275 (2021)
Xu, J. and Luo, Q.: Human action recognition based on mixed gaussian hidden markov model. In: MATEC Web of Conferences, vol. 336, p. 06004. EDP Sciences (2021)
Sargano, A.B., Gu, X., Angelov, P., Habib, Z.: Human action recognition using deep rule-based classifier. Multimedia Tools Appl. 79(41), 30653–30667 (2020)
Ramya, P., Rajeswari, R.: Human action recognition using distance transform and entropy based features. Multimedia Tools Appl. 80(6), 8147–8173 (2020)
Latha, B., Latha, B.M., Manjula, B.K., Sumana, V., Hemalatha, C.H.: Human action recognition using stip evaluation techniques. IOP Conf. Series: Mater. Sci. Eng. 925(1), 12026–12026 (2020)
Abdelbaky, A., Aly, S.: Two-stream spatiotemporal feature fusion for human action recognition. Vis. Comput. 37(7), 1821–1835 (2021)
Berlin, S.J., John, M.: R-STDP based spiking neural network for human action recognition. Appl. Artifi. Intell. 34(9), 656–673 (2020)
Mishra, S.R., Mishra, T.K., Sanyal, G., Sarkar, A., Satapathy, S.C.: Real time human action recognition using triggered frame extraction and a typical CNN heuristic. Pattern Recogn. Lett. 135, 329–336 (2020)
Khan, K., Ali, J., Ahmad, K., Gul, A., Sarwar, G.: 3d head pose estimation through facial features and deep convolutional neural networks. Comput. Mater. Continua 66(2), 1757–1770 (2021)
Rehman, M.U., Ahmed, F., Khan, M.A., Tariq, U., Alfouzan, F.A.: Dynamic hand gesture recognition using 3d-cnn and lstm networks. Comput. Mater. Continua 70(3), 4675–4690 (2022)
Anitha, G., Priya, S.B.: Vision based real time monitoring system for elderly fall event detection using deep learning. Comput. Syst. Sci. Eng. 42(1), 87–103 (2022)
Zou, J., Zhu, N., Ge, B., Hong, D.: Elderly fall detection based on improved SSD algorithm. J. New Media 3(1), 1–10 (2021)
Wang, J., Feng, S., Cheng, Y., Al-Nabhan, N.: Survey on the loss function of deep learning in face recognition. J. Inf. Hiding Privacy Protect. 3(1), 29–45 (2021)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Zhao, X., Hou, F., Su, J., Davis, L. (2022). An Alphapose-Based Pedestrian Fall Detection Algorithm. In: Sun, X., Zhang, X., Xia, Z., Bertino, E. (eds) Artificial Intelligence and Security. ICAIS 2022. Lecture Notes in Computer Science, vol 13338. Springer, Cham. https://doi.org/10.1007/978-3-031-06794-5_52
Download citation
DOI: https://doi.org/10.1007/978-3-031-06794-5_52
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-06793-8
Online ISBN: 978-3-031-06794-5
eBook Packages: Computer ScienceComputer Science (R0)