demiryolu mühendisliği Demiryolu Mühendisliği 2149-1607 2687-2463 Demiryolu Mühendisleri Derneği 10.47072/demiryolu.1918976 Electronic Sensors Photonic and Electro-Optical Devices, Sensors and Systems (Excl. Communications) Elektronik Algılayıcılar Fotonik ve Elektro-Optik Cihazlar, Sensörler ve Sistemler (İletişim Hariç) Otomatik Yolcu Sayma Sistemlerinde Derinlik Bilgisinin Kullanımına Yönelik Kapsamlı Bir İnceleme ve Kavramsal Mimari Çıkarımı A Comprehensive Review and Conceptual Architecture Derivation for the Use of Depth Information in Automated Passenger Counting Systems https://orcid.org/0009-0006-9906-8524 Görgülü Şükrü Albayrak Makine Elektronik San. Tic. A.Ş. https://orcid.org/0000-0003-3410-2227 Adıyaman Soner Albayrak Makine Elektronik San. Tic. A.Ş. https://orcid.org/0000-0002-5801-8292 Öden Seray Albayrak Makine Elektronik San. Tic. A.Ş. https://orcid.org/0000-0003-3290-914X Kırımça Necim Albayrak Makine Elektronik San. Tic. A.Ş. 05 04 2026 24 03 30 2026 05 04 2026 Copyright © 2014, Demiryolu Mühendisliği 2014 Demiryolu Mühendisliği

Toplu taşıma sistemlerinde yolcu sayısı, hizmet planlaması ve operasyonel verimlilik açısından kritik bir girdidir. Ancak yolcu yoğunluğu; saat, hava koşulları ve yerel etkinlikler gibi faktörlere bağlı olarak sürekli değişmektedir. Geleneksel biletleme ve turnike tabanlı yöntemler, dolaylı veri üretmeleri ve insan faktörüne bağlı olmaları nedeniyle özellikle yoğun ortamlarda doğruluk sorunları yaratmaktadır. Bu nedenle otomatik yolcu sayma sistemleri önem kazanmıştır. Özellikle derinlik verisi kullanan yaklaşımlar son yıllarda öne çıkmıştır. Bu çalışmada, yolcu sayımında derinlik bilgisinin kullanımı sistematik olarak incelenmektedir. Renk+derinlik, stereo görüş ve uçuş süresi (Time of Flight - ToF) sensörlerinin otomatik yolcu sayma uygulamalarındaki rolleri incelenmiştir. Üstten bakış konfigürasyonunun örtülmeye (occlusion) karşı dayanıklı olması ve yön ayrıştırma başarımı sayesinde baskın mimari haline geldiği görülmüştür. Literatür, sensör tercihlerinin stereo sistemlerden ToF teknolojilerine evrildiğini; algoritmik olarak ise klasik segmentasyondan çok aşamalı tespit-takip ve öğrenme tabanlı hibrit modellere geçiş yaşandığını göstermektedir. Modern sistemlerde hata oranlarının %1–3 seviyelerine yaklaştığı ve 30-60 kare hızı ile gerçek zamanlılık hedeflerinin sağlandığı görülmektedir. Ayrıca mevcut literatür tarama çalışmalarında otomatik yolcu sayma sistemlerinin bileşenleri tekil olarak incelenmekte ve bileşenlerin sistem seviyesinde nasıl bir araya getirileceğine dair bütüncül bir çerçeve sunulmamaktadır. Bu nedenle, literatürdeki ortak tasarım desenlerinden hareketle, derinlik tabanlı otomatik yolcu sayma sistemleri için hem modüler hem de bütüncül bir referans mimari çıkarılmıştır. Bu mimari, algılama, ön işleme, insan tespiti, çoklu nesne takibi ve yön tabanlı sayım katmanlarını birleştirmektedir. Çalışmada, demiryolu toplu taşıma sistemlerinde otomatik yolcu sayımı için derinlik bilgisinin yalnızca sensör seçiminde değil, aynı zamanda sistem mimarisinde ve algoritma tasarımında önemli bir unsur olduğu sonucuna varılmıştır.

Passenger volume is a key input for service planning and operations in public transport systems. However, it varies dynamically with factors such as time of day, weather, and local events, making accurate and real-time estimation essential for both efficiency and passenger comfort. Traditional ticketing and turnstile-based methods rely on indirect data and human factors, often leading to inaccuracies in crowded and dynamic environments. As a result, automated passenger counting (APC) systems have gained prominence, particularly those leveraging depth information. This study examines the use of depth-based approaches, focusing on RGB-D, Stereo Vision, and Time-of-Flight (ToF) sensors. Findings indicate that top-down configurations have become dominant due to their robustness against occlusion and effectiveness in direction detection. The literature also highlights a technological shift from stereo vision to ToF sensors, alongside a transition from classical segmentation methods to multi-stage detection, tracking, and learning-based hybrid models. Modern APC systems achieve error rates as low as 1–3% while maintaining real-time performance of 30–60 frames/second. Existing literature reviews examine the components of APC systems individually, and a holistic framework for integrating them at the system level is lacking. Therefore, based on common patterns in the literature, a modular reference architecture was derived for depth-based APC systems, integrating sensing, preprocessing, human detection, multi-object tracking, and direction-based counting. The study concludes that depth information is not only a topic for sensor choice but a key factor shaping system architecture and algorithmic design.

 Otomatik Yolcu Sayma Derinlik Kamerası Time-Of-Flight Üstten Bakış Yolcu Tespiti Takip Automatic Passenger Counting Depth Camera Time-Of-Flight Overhead Passenger Detection Tracking T.C. Ulaştırma ve Altyapı Bakanlığı, Ulaştırma, Denizcilik ve Haberleşme Araştırmaları Merkezi Başkanlığı (UDHAM) 2023.UT.01.01.004 [1] J. Salvi, S. Fernandez, T. Pribanic, and X. Llado, “A state of the art in structured light patterns for surface profilometry,” Pattern Recognition, vol. 43, no. 8, pp. 2666–2680, 2010, doi: 10.1016/j.patcog.2010.03.004 [2] Z. Zhang, “Microsoft kinect sensor and its effect,” IEEE Multimedia, vol. 19, no. 2, pp. 4–10, 2012, doi: 10.1109/MMUL.2012.24 [3] D. Scharstein and R. Szeliski, “A taxonomy and evaluation of dense two-frame stereo correspond-ence algorithms,” International Journal of Computer Vision, vol. 47, no. 1–3, pp. 7–42, 2002, doi: 10.1023/A:1014573219977 [4] R. Hartley and A. Zisserman, Multiple view geometry in computer vision. Cambridge, U.K.: Cam-bridge University Press, 2004. [5] R. Lange and P. Seitz, “Solid-state time-of-flight range camera,” IEEE Journal of Quantum Electron-ics, vol. 37, no. 3, pp. 390–397, 2001. [6] A. Kolb, E. Barth, R. Koch, and R. Larsen, “Time-of-flight sensors in computer graphics,” Computer Graphics Forum, vol. 29, no. 1, pp. 141–159, 2010, doi: 10.1111/j.1467-8659.2009.01583.x [7] H. Zhang, C. Reardon, and L. E. Parker, “Real-time multiple human perception with color-depth cameras on a mobile robot,” IEEE Transactions on Cybernetics, vol. 43, no. 5, pp. 1429–1441, 2013, doi: 10.1109/TCYB.2013.2275291 [8] O. H. Jafari, D. Mitzel, and B. Leibe, “Real-time RGB-D based people detection and tracking for mobile robots and head-worn cameras,” in Proc. IEEE Int. Conf. Robotics and Automation (ICRA), 2014, pp. 5636–5643, doi: 10.1109/ICRA.2014.6907688 [9] J. Liu, Y. Liu, Y. Cui, and Y. Q. Chen, “Real-time human detection and tracking in complex envi-ronments using single RGB-D camera,” in Proc. IEEE Int. Conf. Image Processing, 2013, pp. 3088–3092, doi: 10.1109/ICIP.2013.6738636 [10] J. Liu, Y. Liu, G. Zhang, P. Zhu, and Y. Q. Chen, “Detecting and tracking people in real time with RGB-D camera,” Pattern Recognition Letters, vol. 53, pp. 16–23, 2015, doi: 10.1016/j.patrec.2014.09.013 [11] J. Liu, G. Zhang, Y. Liu, L. Tian, and Y. Q. Chen, “An ultra-fast human detection method for color-depth camera,” Journal of Visual Communication and Image Representation, vol. 31, pp. 177–185, 2015, doi: 10.1016/j.jvcir.2015.06.014 [12] Y. Xiao, V. R. Kamat, and C. C. Menassa, “Human tracking from single RGB-D camera using online learning,” Image and Vision Computing, vol. 88, pp. 67–75, 2019, doi: 10.1016/j.imavis.2019.05.003 [13] Y. Yao, X. Zhang, Y. Liang, X. Zhang, F. Shen, and J. Zhao, “A real-time pedestrian counting system based on RGB-D,” in Proc. 12th Int. Conf. Advanced Computational Intelligence (ICACI), 2020, pp. 110–117, doi: 10.1109/ICACI49185.2020.9177816 [14] M. Gochoo, S. A. Rizwan, Y. Y. Ghadi, A. Jalal and K. Kim, “A systematic deep learning based over-head tracking and counting system using RGB-D remote cameras,” Applied Sciences, vol. 11, no. 12, pp. 5503, 2021, doi: 10.3390/app11125503 [15] L. Tian, M. Li, Y. Hao, J. Liu, G. Zhang, and Y. Q. Chen, “Robust 3-D human detection in complex environments with a depth camera,” IEEE Transactions on Multimedia, vol. 20, no. 9, pp. 2249–2261, 2018, doi: 10.1109/TMM.2018.2803526 [16] L. Xia, C.-C. Chen, and J. K. Aggarwal, “Human detection using depth information by kinect,” in CVPR 2011 Workshops, 2011, pp. 15–22, doi: 10.1109/CVPRW.2011.5981811 [17] F. Galčík and R. Gargalík, “Real-time depth map based people counting,” in International Confer-ence on Advanced Concepts for Intelligent Vision Systems, 2013, pp. 330–341. [18] L. Zhu and K.-H. Wong, “Human tracking and counting using the kinect range sensor based on ada-boost and kalman filter,” in International Symposium on Visual Computing (ISVC), 2013, pp. 582–591. [19] M. H. Khan, K. Shirahama, M.S. Farid and M. Grzegorzek, “Multiple human detection in depth im-ages,” in IEEE 18th Int. Workshop Multimedia Signal Processing (MMSP), 2016, pp. 1–6, doi: 10.1109/MMSP.2016.7813385 [20] Z. Qiuyu, T. Li, J. Yiping, and D. Wei-jun, “A novel approach of counting people based on stereovi-sion and DSP,” in The 2nd. Int. Conf. on Computer and Automation Engineering (ICCAE), 2010, pp. 81–84, doi: 10.1109/ICCAE.2010.5451996 [21] N. Bernini, L. Bombini, M. Buzzoni, P. Cerri, and P. Grisleri, “An embedded system for counting passengers in public transportation vehicles,” in IEEE/ASME 10th Int. Conf. Mechatronic and Em-bedded Systems and Applications (MESA), 2014, pp. 1–6, doi: 10.1109/MESA.2014.6935562 [22] T. Yahiaoui, L. Khoudour, and C. Meurie, “Real-time passenger counting in buses using dense ste-reovision,” Journal of Electronic Imaging, vol. 19, no. 3, 2010, doi: 10.1117/1.3455989 [23] M. Rauter, “Reliable human detection and tracking in top-view depth images,” in Proc. IEEE CVPR Workshops, 2013, pp. 529–534. [24] D. Nurseitov, K. Bostanbekov, N. Toiganbayeva, A. Zhalgas, D. Yedilkhan, and B. Amirgaliyev, "Vision-based people counting and tracking for urban environments," Journal of Imaging, vol. 12, no. 1, pp. 27, 2026, doi: 10.3390/jimaging12010027. [25] D. Klauser, G. Bärwolff, and H. Schwandt, “A TOF-based automatic passenger counting approach in public transportation systems,” AIP Conference Proceedings, vol. 1648, no. 1, pp. 850113, 2015, doi: 10.1063/1.4913168 [26] A. Bevilacqua, L. Di Stefano, and P. Azzari, “People tracking using a time-of-flight depth sensor,” in Proc. IEEE Int. Conf. on Video and Signal Based Surveillance, 2006, pp. 89–89, doi: 10.1109/AVSS.2006.92 [27] R. Tanner, M. Studer, A. Zanoli, and A. Hartmann, “People detection and tracking with tof sensor,” in IEEE Fifth International Conference on Advanced Video and Signal Based Surveillance, 2008, pp. 356–361, doi: 10.1109/AVSS.2008.18 [28] S. Ikemura and H. Fujiyoshi, “Human detection by haar-like filtering using depth information,” in Proc. of the 21st ICPR, 2012, pp. 813–816. [29] C. Stahlschmidt, A. Gavriilidis, J. Velten and A. Kummert, “Applications for a people detection and tracking algorithm using a time-of-flight camera,” Multimedia Tools and Applications, vol. 75, no. 17, pp. 10769–10786, 2016. [30] C.A. Luna Vazquez, C. Losada Gutierrez, D. Fuentes Jimenez, A. Fernandez Rincon, M. R. Mazo Quintas, and J. Macias Guarasa “Robust people detection using depth information from an overhead time-of-flight camera,” Expert Systems with Applications, vol. 71, pp. 240–256, 2017, doi: 10.1016/j.eswa.2016.11.019 [31] A. Fernandez-Rincon, D. Fuentes-Jimenez, C. Losada-Gutierrez, M. Marron-Romera, C.A. Luna, J. Macias-Guarasa and M. Mazo, “Robust people detection and tracking from an overhead time-of-flight camera,” in International Conference on Computer Vision Theory and Applications, 2017, pp. 556–564. [32] W. Wang, P. Liu, R. Ying, J. Wang, J. Qian, J. Jia, J. Gao, “A high-computational efficiency human detection and flow estimation method based on tof measurements,” Sensors, vol. 19, no. 3, pp. 729, 2019, doi: 10.3390/s19030729 [33] F. Zoghlami, O.K. Sen, H. Heinrich, G. Schneider, E. Ercelik, A. Knoll, T. Villmann, “ToF/radar early feature-based fusion system for human detection and tracking,” in 22nd IEEE ICIT, 2021, pp. 942–949, doi: 10.1109/ICIT46573.2021.9453703 [34] W. Rahmaniar, W. Wang, C. Chiu and N. Hakim, “Real-time bi-directional people counting using an RGB-D camera,” Sensor Review, vol. 41, no. 4, pp. 341-349, 2021, doi: 10.1108/SR-12-2020-0301 [35] C. Stahlschmidt, A. Gavriilidis, J. Velten, and A. Kummert, “People detection and tracking from a top-view position using a time-of-flight camera,” in MCSS, 2013, pp. 213–223. [36] L. Tamas and A. Cozma, “Embedded real-time people detection and tracking with time-of-flight camera,” in Real-Time Image Processing and Deep Learning, vol. 11736, pp. 65-70, 2021, doi: 10.1117/12.2586057 [37] T.-E. Tseng, A.-S. Liu, P.-H. Hsiao, C.-M. Huang, and L.-C. Fu, “Real-time people detection and tracking for indoor surveillance using multiple top-view depth cameras,” in. IEEE/RSJ IROS, 2014, pp. 4077–4082, doi: 10.1109/IROS.2014.6943136 [38] S. T. Dharmarajan, Y. Zhang, B. Li, M. Saada, H. Cai, Q. Meng, Q. Zhu, Q. Han, “Development of a depth imaging-based passenger counting sensor for public transportation,” in IEEE RAICS, 2025, pp. 192–200, doi: 10.1109/RAICS66191.2025.11332612 [39] S. Thandassery, Y. Zhang, B. Li, M. Saada, H. Cai, Q. Meng, Q. Zhu, Q. Han “Smart railway passen-ger counting and information systems powered by real-time embedded AI and computer vision,” in IEEE ICVES, 2025, pp. 453–459, doi: 10.1109/ICVES65691.2025.11376469