Analyzing and Recognizing Pedestrian Motion Using 3D Sensor Network and Machine Learning
How to analyze and recognize pedestrian movements is an important issue dependent on motion capture devices. In our work, we used two types of popular 3D sensors such as 3D depth sensor and 3D motion sensor to construct a sensor network for tacking motion of target because of their convenience and low cost. In this paper, we first describe how to get data from the sensor network and how to process raw data. Next, we provide algorithms for applying machine learning to the analysis and recognition of human motions. Finally, we give some evaluation experimental results.
- 1.Sun, N., Murakami, S.: Human motion modeling from complementary skeleton joints of multiple kinects. In: Proceedings of the 2018 International Conference on Mathematics, Modelling, Simulation and Algorithms. Advances in Intelligent Systems Research, vol. 159, pp. 131–135 (2018)Google Scholar
- 2.Sun, N., Sakai, Y.: New approaches to human gait simulation using motion sensors. In: 31st International Conference on Advanced Information Networking and Applications Workshops (WAINA) (2017). 16901760Google Scholar
- 3.Sun, N., Tsuruoka, T.: Pedestrian action recognition using motion sensor and k-nn classifier. In: Proceedings of 2nd International Conference on Artificial Intelligence: Technologies and Applications. Advances in Intelligent Systems Research, vol. 146, pp. 1–4 (2018)Google Scholar
- 5.Foley, J.D., Cardelli, L., van Dam, A., Feiner, S.K., Hughess, J.F.: Computer Graphics: Principles and Practice, 3rd edn., pp. 263–286, 299–320. Addison-Wesley, New York (2015)Google Scholar