The choice of camera mounting height and pedestrian tracking strategy – by body or head bounding boxes – is examined as a function of flow density. The mounting height and the observed body part govern mutual occlusions in the image and thereby cap the achievable quality of multi-object tracking before any tracker is chosen. Quality is measured by an integral higher-order tracking metric, and for each strategy the minimum mounting height at which this quality reaches a target level is estimated. The estimation uses synthetic video sequences that reproduce a real pedestrian flow under controllable acquisition geometry that is unattainable in field recordings with a predetermined viewpoint. The computational experiment spans six mounting heights from 2 to 12 m and a density range from free flow to dense flow. For body tracking the required height grows monotonically with density and exceeds the practical limit in dense regimes, whereas head tracking attains the same quality at a height of no more than 6 m. The obtained dependence is applicable at the design stage of video surveillance systems for matching the mounting height and observation strategy to the expected pedestrian flow density.
1. Milan A., Leal-Taixé L., Reid I., Roth S., Schindler K. MOT16: A benchmark for multi-object tracking. arXiv. URL: https://doi.org/10.48550/arXiv.1603.00831 [Accessed 15th May 2026].
2. Luiten J., Os̆ep A., Dendorfer P., et al. HOTA: A higher order metric for evaluating multi-object tracking. International Journal of Computer Vision. 2021;129(11):548–578. https://doi.org/10.1007/s11263-020-01375-2
3. Erdem U.M., Sclaroff S. Automated camera layout to satisfy task-specific and floor plan-specific coverage requirements. Computer Vision and Image Understanding. 2006;103(3):156–169. https://doi.org/10.1016/j.cviu.2006.06.005
4. Volkhonskiy V.V., Kovalevskiy V.A. Method for optimization of camera installation parameters for video monitoring of arbitrary surveillance zone. Scientific and Technical Journal of Information Technologies, Mechanics and Optics. 2023;23(5):927–934. (In Russ.). https://doi.org/10.17586/2226-1494-2023-23-5-927-934
5. Limanta F., Uto K., Shinoda K. CAMOT: Camera angle-aware multi-object tracking. In: Proceedings of the 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 03–08 January 2024, Waikoloa, HI, USA. IEEE; 2024. P. 6465–6474. https://doi.org/10.1109/WACV57701.2024.00635
6. Sundararaman R., De Almeida Braga C., Marchand E., Pettré J. Tracking pedestrian heads in dense crowd. In: Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 20–25 June 2021, Nashville, TN, USA. IEEE; 2021. P. 3864–3874. https://doi.org/10.1109/CVPR46437.2021.00386
7. Zhang Y., Chen H., Lai Zh., Zhang Z., Yuan D. Handling heavy occlusion in dense crowd tracking by focusing on the heads. In: AI 2023: Advances in Artificial Intelligence: 36th Australasian Joint Conference on Artificial Intelligence: Part I, 28 November – 01 December 2023, Brisbane, QLD, Australia. Singapore: Springer; 2023. P. 79–90. https://doi.org/10.1007/978-981-99-8388-9_7
8. Helbing D., Molnár P. Social force model for pedestrian dynamics. Physical Review E. 1995;51(5):4282–4286. https://doi.org/10.1103/PhysRevE.51.4282
9. Dendorfer P., Rezatofighi H., Milan A., et al. MOT20: A benchmark for multi object tracking in crowded scenes. arXiv. URL: https://doi.org/10.48550/arXiv.2003.09003 [Accessed 15th May 2026].
10. Simsek F.E., Cigla C., Kayabol K. SOMPT22: A surveillance oriented multi-pedestrian tracking dataset. arXiv. URL: https://doi.org/10.48550/arXiv.2208.02580 [Accessed 15th May 2026].
11. NCD Risk Factor Collaboration (NCD-RisC). A century of trends in adult human height. eLife. 2016;5:e13410. https://doi.org/10.7554/eLife.13410
12. Mather G. Head-body ratio as a visual cue for stature in people and sculptural art. Perception. 2010;39(10):1390–1395. https://doi.org/10.1068/p6737
13. Cao J., Pang J., Weng X., Khirodkar R., Kitani K. Observation-centric SORT: Rethinking SORT for robust multi-object tracking. In: Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 17–24 June 2023, Vancouver, BC, Canada. IEEE; 2023. P. 9686–9696. https://doi.org/10.1109/CVPR52729.2023.00934
14. Bewley A., Ge Z., Ott L., Ramos F., Upcroft B. Simple online and realtime tracking. In: Proceedings of the 2016 IEEE International Conference on Image Processing (ICIP), 25–28 September 2016, Phoenix, AZ, USA. IEEE; 2016. P. 3464–3468. https://doi.org/10.1109/ICIP.2016.7533003
15. Zhang Y., Sun P., Jiang Y., et al. ByteTrack: Multi-object tracking by associating every detection box. In: Computer Vision – ECCV 2022: 17th European Conference: Part XXII, 23–27 October 2022, Tel Aviv, Israel. Cham: Springer; 2022. P. 1–21. https://doi.org/10.1007/978-3-031-20047-2_1
Nikashov Ilya Alekseevich
Vladimir State University named after Alexander Grigorievich and Nikolai Grigorievich Stoletov
Vladimir, Russian Federation
