Lost Centroid ID in CCTV Analytics for Slow Moving Object

Ulva Elviani, Hanavi Hanavi, Fadhil Hidayat

Abstract


Accuracy is the most common problem experienced by traditional CCTV, especially people counting in locations where people are in the camera's field of view for a long time. This study proposes a method to address the lost centroid ID problem through a time delay mechanism in the object detection and tracking path using YOLO. This method can reduce the number of duplications and improve overall data reliability by maintaining object identities in predetermined frames, even when objects are temporarily undetected. Experiments were conducted on 100 CCTV video recording files in the Bandung train station area. The results showed that in locations in the slow-moving object category, such as waiting rooms, the accuracy increased significantly from 42% to 72% by applying a maxdisappeared threshold value of 200 frames. While at fast-moving object locations, a threshold of 40 frames increased the accuracy from 83% to 94%. This approach improves the performance of the people counting forecasting model for a more reliable surveillance system both statically and dynamically.


Keywords


CCTV Analytics, Centroid ID, People Counting, Slow-moving object, Video Surveillance.

Full Text:

PDF

References


Barthélemy, J., Verstaevel, N., Forehead, H., & Perez, P. (2019). Edge-computing video analytics for real-time traffic monitoring in a smart city. Sensors (Switzerland), 19(9), 16–17. https://doi.org/10.3390/s19092048

Bisht, A., Mishra, U., & Saraf, S. (2024). People Analytics Using Deep Learning. 2nd IEEE International Conference on Advances in Information Technology, ICAIT 2024 - Proceedings. https://doi.org/10.1109/ICAIT61638.2024.10690501

Choi, H., Fujimoto, M., Matsui, T., Misaki, S., & Yasumoto, K. (2022). Wi-CaL: WiFi Sensing and Machine Learning Based Device-Free Crowd Counting and Localization. IEEE Access, 10, 24395–24410. https://doi.org/10.1109/ACCESS.2022.3155812

Chua, D. E. P., Recto, K. H. A., & Mayuga, G. P. T. (2023). Real-Time Human Detection and Tracking System: A Novel Comparative Study of Centroid Tracking, Single Shot Detection and YOLO Algorithms. 2023 1st International Conference on Advanced Engineering and Technologies, ICONNIC 2023 - Proceeding, 97–102. https://doi.org/10.1109/ICONNIC59854.2023.10467636

Ibrahim, O. T., Gomaa, W., & Youssef, M. (2019). CrossCount: A Deep Learning System for Device-Free Human Counting Using WiFi. IEEE Sensors Journal, 19(21), 9921–9928. https://doi.org/10.1109/JSEN.2019.2928502

Kouyoumdjieva, S. T., Danielis, P., & Karlsson, G. (2020). Survey of Non-Image-Based Approaches for Counting People. IEEE Communications Surveys and Tutorials, 22(2), 1305–1336. https://doi.org/10.1109/COMST.2019.2902824

Li, Y., Qu, L., Cai, G., Cheng, G., Qian, L., Dou, Y., Yao, F., & Wang, S. (2023). Video Object Counting with Scene-Aware Multi-Object Tracking. Journal of Database Management, 34(3). https://doi.org/10.4018/JDM.321553

Liu, Z., Chen, Y., Chen, B., Zhu, L., Wu, D., & Shen, G. (2019). Crowd Counting Method Based on Convolutional Neural Network With Global Density Feature. IEEE Access, 7, 88789–88798. https://doi.org/10.1109/ACCESS.2019.2926881

Meng, C., Zhang, Y., & Hou, Q. (2023). Person Re-identification Based on Centroids Triplet Loss. ACM International Conference Proceeding Series, 306–311. https://doi.org/10.1145/3641584.3641630

Narayanan, S. J., Perumal, B., Saman, S., & Singh, A. P. (2020). Deep learning for person re-identification in surveillance videos. Studies in Computational Intelligence, 865, 263 – 297. https://doi.org/10.1007/978-3-030-31760-7_9

Nguyen, K.-D., Nguyen, H. H., Le, T.-N., Yamagishi, J., & Echizen, I. (2024). Analysis of Fine-Grained Counting Methods for Masked Face Counting: A Comparative Study. IEEE Access, 12, 27426–27443. https://doi.org/10.1109/ACCESS.2024.3367593

Putra, M. P. K., & Wahyono. (2021). A Novel Method for Handling Partial Occlusion on Person Re-identification using Partial Siamese Network. International Journal of Advanced Computer Science and Applications, 12(7), 313 – 321. https://doi.org/10.14569/IJACSA.2021.0120735

Raja, Y., Gong, S., & Xiang, T. (2011). User-assisted visual search and tracking across distributed multi-camera networks. Proceedings of SPIE - The International Society for Optical Engineering, 8189. https://doi.org/10.1117/12.897673

Shakil A.*, S., & Kureshi, D. A. K. (2020). Object Detection and Tracking using YOLO v3 Framework for Increased Resolution Video. International Journal of Innovative Technology and Exploring Engineering, 9(6), 118–125. https://doi.org/10.35940/ijitee.e3038.049620

Ullah, R., Hayat, H., Siddiqui, A. A., Siddiqui, U. A., Khan, J., Ullah, F., Hassan, S., Hasan, L., Albattah, W., Islam, M., & Karami, G. M. (2022). A Real-Time Framework for Human Face Detection and Recognition in CCTV Images. Mathematical Problems in Engineering, 2022. https://doi.org/10.1155/2022/3276704

Violatto, G., & Pandharipande, A. (2020). Anomaly Classification in People Counting and Occupancy Sensor Systems. IEEE Sensors Journal, 20(12), 6573–6581. https://doi.org/10.1109/JSEN.2020.2976547

Zhang, H., Zhou, M., Sun, H., Zhao, G., Qi, J., Wang, J., & Esmaiel, H. (2021). Que-Fi: A Wi-Fi Deep-Learning-Based Queuing People Counting. IEEE Systems Journal, 15(2), 2926–2937. https://doi.org/10.1109/JSYST.2020.2994062




DOI: https://doi.org/10.17509/coelite.v4i2.85752

Refbacks

  • There are currently no refbacks.


Journal of Computer Engineering, Electronics and Information Technology (COELITE)


is published by UNIVERSITAS PENDIDIKAN INDONESIA (UPI),
and managed by Department of Computer Enginering.
Jl. Dr. Setiabudi No.229, Kota Bandung, Indonesia - 40154
email: coelite@upi.edu
e-ISSN: 2829-4149
p-ISSN: 2829-4157