Publicación: Uso de videoanalı́tica y modelos matemáticos para la detección y seguimiento de objetos para apoyar la gestión vial en la ciudad de Bogotá
| dc.contributor.advisor | Luque Zabala, Carolina María | |
| dc.contributor.author | Moreno Bedoya, David Leonardo | |
| dc.contributor.researchgroup | Ciencia, tecnología e innovación::Ciencias básicas Miguel Ángel González Curbelo Categoría A COL0036553 | |
| dc.creator.id | 79891338 | |
| dc.date.accessioned | 2026-07-10T21:50:48Z | |
| dc.date.issued | 2026-06-13 | |
| dc.description.abstract | Bogotá enfrenta una alta accidentalidad vial, con fallecidos que aumentan progresivamente (508 en 2021, 665 en 2024), siendo el exceso de velocidad uno de los principales factores contribuyentes. Los sistemas actuales de videoanalítica de la Secretaría de Movilidad alcanzan una precisión cercana al 80 %, insuficiente para los modelos de calibración de congestión que requieren más del 85 %. Este proyecto propone un sistema de seguimiento multi-objeto (MOT) que integra detección YOLO con un Filtro de Kalman de 8 dimensiones, velocidades estimadas mediante flujo óptico de Lucas-Kanade y descriptores de apariencia OSNet. La asociación óptima se resuelve mediante el algoritmo Húngaro, combinando IoU y similitud coseno. La metodología cuantitativa empleó tres datasets: Waymo Open Dataset (990 frames, cámara móvil), MOT17 (5316 frames, cámara estática) y un dataset local de Bogotá (500 frames, cámara fija), con un estudio de ablación de 14 configuraciones. Los resultados muestran que el tracker propuesto alcanza 85.26 % MOTA y el mejor IDF1 (87.64 %) en el dataset de Bogotá, superando marginalmente el umbral del 85 %, aunque con MOTA inferior a ByteTrack (89.15 %) y SORT (89.63 %). Su principal ventaja reside en la preservación de identidades (32 ID switches vs 103 de SORT), dimensión crítica para estimación de trayectorias. El estudio de ablación revela que el ajuste de hiperparámetros (min hits=1: +9.14 % MOTA) tiene mayor impacto que los componentes algorítmicos (flujo óptico: +3.45 %, embeddings: +2.76 %). Se concluye que el sistema ofrece evidencia preliminar de viabilidad para la Secretaría de Movilidad en el escenario evaluado, con configuraciones adaptables según las necesidades de procesamiento. La validación en condiciones más diversas es necesaria antes de un despliegue operativo. | spa |
| dc.description.abstract | Bogotá faces a high rate of traffic accidents, with fatalities steadily increasing (508 in 2021, 665 in 2024), with speeding being one of the main contributing factors. The current video analytics systems of the Mobility Secretariat achieve an accuracy of approximately 80%, insufficient for congestion calibration models that require over 85%. This project proposes a multi-object tracking (MOT) system that integrates YOLO detection with an 8-dimensional Kalman filter, speeds estimated using Lucas-Kanade optical flow, and OSNet appearance descriptors. Optimal association is achieved using the Hungarian algorithm, combining IoU and cosine similarity. The quantitative methodology employed three datasets: Waymo Open Dataset (990 frames, mobile camera), MOT17 (5316 frames, static camera), and a local Bogotá dataset (500 frames, fixed camera), with an ablation study of 14 configurations. The results show that the proposed tracker achieves 85.26% MOTA and the best IDF1 (87.64%) in the Bogotá dataset, marginally exceeding the 85% threshold, although with a lower MOTA than ByteTrack (89.15%) and SORT (89.63%). Its main advantage lies in identity preservation (32 ID switches vs. 103 for SORT), a critical dimension for path estimation. The ablation study reveals that hyperparameter adjustment (min hits=1: +9.14% MOTA) has a greater impact than algorithmic components (optical flow: +3.45%, embeddings: +2.76%). It is concluded that the system offers preliminary evidence of viability for the Mobility Secretariat in the evaluated scenario, with adaptable configurations according to processing needs. Validation under more diverse conditions is necessary before operational deployment. | eng |
| dc.description.degreelevel | Maestría | spa |
| dc.description.degreename | Magíster en Ciencias de Datos | spa |
| dc.description.researcharea | CIENCIA, TECNOLOGÍA E INNOVACIÓN::CIENCIAS BASICAS MIGUEL ANGEL GONZALEZ CURBELO Categoría A COL0036553::Matemáticas aplicadas | |
| dc.format | ||
| dc.format.extent | 65 páginas | |
| dc.format.medium | Recurso electrónico | spa |
| dc.format.mimetype | application/pdf | |
| dc.identifier.instname | instname:Universidad Ean | spa |
| dc.identifier.local | BDM-MGP | |
| dc.identifier.reponame | reponame:Repositorio Institucional Biblioteca Digital Minerva | spa |
| dc.identifier.repourl | repourl:https://repository.ean.edu.co/ | |
| dc.identifier.uri | https://hdl.handle.net/10882/19363 | |
| dc.publisher.faculty | Facultad de Ingeniería | spa |
| dc.publisher.program | Maestría en Ciencias de Datos | spa |
| dc.relation.references | Aharon, N., Orfaig, R., & Bobrovsky, B.-Z. (2022). BoT-SORT: Robust Associations Multi-Pedestrian Tracking. arXiv preprint arXiv:2206.14651. Bernardin, K., & Stiefelhagen, R. (2008). Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics. EURASIP Journal on Image and Video Processing, 2008, 1-10. https://doi.org/10.1155/2008/246309 Bewley, A., Ge, Z., Ott, L., Ramos, F., & Upcroft, B. (2016). Simple Online and Realtime Tracking. IEEE International Conference on Image Processing (ICIP), 3464-3468. https://doi.org/10.1109/ICIP.2016.7533003 Bloomberg Philanthropies. (2004). Initiative for Global Road Safety (Reporte sobre seguridad vial global). Bloomberg Philanthropies. Cao, J., Weng, X., Khirodkar, R., Pang, J., & Kitani, K. (2023). Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 9686-9696. Ciaparrone, G., Luque Sánchez, F., Tabik, S., Troiano, L., Tagliaferri, R., & Herrera, F. (2020). Deep Learning in Video Multi-Object Tracking: A Survey. Neurocomputing, 381, 61-88. https://doi.org/10.1016/j.neucom.2019.11.023 Dendorfer, P., Osep, A., Milan, A., Schindler, K., Cremers, D., Reid, I., Roth, S., & Leal-Taixé, L. (2021). MOTChallenge: A Benchmark for Single-Camera Multiple Target Tracking. International Journal of Computer Vision, 129, 845-881. https://doi.org/10.1007/s11263-020-01393-0 Du, Y., Zhao, Z., Song, Y., Zhao, Y., Su, F., Gong, T., & Meng, H. (2023). StrongSORT: Make DeepSORT Great Again. IEEE Transactions on Multimedia, 25, 8725-8737. https://doi.org/10.1109/TMM.2023.3240881 Farnebäck, G. (2003). Two-Frame Motion Estimation Based on Polynomial Expansion. Proceedings of the 13th Scandinavian Conference on Image Analysis (SCIA), 2749, 363-370. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep Residual Learning for Image Recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770-778. Hernández Sampieri, R., & Mendoza Torres, C. P. (2018). Metodología de la investigación: las rutas cuantitativa, cualitativa y mixta (7.a ed.). McGraw-Hill. Jocher, G., Chaurasia, A., & Qiu, J. (2023). Ultralytics YOLOv8. https://github.com/ultralytics/ultralytics Kalman, R. E. (1960). A New Approach to Linear Filtering and Prediction Problems. Transactions of the ASME–Journal of Basic Engineering, 82(Series D), 35-45. Kuhn, H. W. (1955). The Hungarian Method for the Assignment Problem. Naval Research Logistics Quarterly, 2(1-2), 83-97. https://doi.org/10.1002/nav.3800020109 Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014). Microsoft COCO: Common Objects in Context. European Conference on Computer Vision (ECCV), 740-755. https://doi.org/10.1007/978-3-319-10602-1_48 Lucas, B. D., & Kanade, T. (1981). An Iterative Image Registration Technique with an Application to Stereo Vision. Proceedings of the 7th International Joint Conference on Artificial Intelligence, 674-679. Luiten, J., Osep, A., Dendorfer, P., Torr, P., Geiger, A., Leal-Taixé, L., & Leibe, B. (2021). HOTA: A Higher Order Metric for Evaluating Multi-Object Tracking. International Journal of Computer Vision, 129(2), 548-578. https://doi.org/10.1007/s11263-020-01375-2 Luo, W., Xing, J., Milan, A., Zhang, X., Liu, W., & Kim, T.-K. (2021). Multiple Object Tracking: A Literature Review. Artificial Intelligence, 293, 103448. https://doi.org/10.1016/j.artint.2020.103448 Maggiolino, G., Ahmad, A., Cao, J., & Kitani, K. (2023). Deep OC-SORT: Multi-Pedestrian Tracking by Adaptive Re-Identification. Proceedings of the IEEE International Conference on Image Processing (ICIP), 3025-3029. Milan, A., Leal-Taixe, L., Reid, I., Roth, S., & Schindler, K. (2016). MOT16: A Benchmark for Multi-Object Tracking. arXiv preprint arXiv:1603.00831. Moreno Bedoya, D. L. (2026). Understanding Motion Models in Multi-Object Tracking: A Rigorous Analysis from SORT to OC-SORT with Theory-Guided Tracker Selection [Manuscript under review]. Munkres, J. (1957). Algorithms for the Assignment and Transportation Problems. Journal of the Society for Industrial and Applied Mathematics, 5(1), 32-38. Padilla, R., Netto, S. L., & da Silva, E. A. B. (2020). A Survey on Performance Metrics for Object-Detection Algorithms. Proceedings of the International Conference on Systems, Signals and Image Processing (IWSSIP), 237-242. Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. (2016). You Only Look Once: Unified, Real-Time Object Detection. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 779-788. Rezatofighi, H., Tsoi, N., Gwak, J., Sadeghian, A., Reid, I., & Savarese, S. (2019). Generalized Intersection over Union: A Metric and A Loss for Bounding Box Regression. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 658-666. Ristani, E., Solera, F., Zou, R., Cucchiara, R., & Tomasi, C. (2016). Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking. European Conference on Computer Vision (ECCV) Workshops, 17-35. https://doi.org/10.1007/978-3-319-48881-3_2 Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L.-C. (2018). MobileNetV2: Inverted Residuals and Linear Bottlenecks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 4510-4520. Secretaría Distrital de Movilidad. (2023). Plan Distrital de Desarrollo 2024-2027: Bogotá Camina Segura. Sun, P., Kretzschmar, H., Dotiwalla, X., Chouard, A., Patnaik, V., Tsui, P., Guo, J., Zhou, Y., Chai, Y., Caine, B., Vasudevan, V., Han, W., Ngiam, J., Zhao, H., Timofeev, A., Ettinger, S., Krivokon, M., Gao, A., Joshi, A., ... Anguelov, D. (2020). Scalability in Perception for Autonomous Driving: Waymo Open Dataset. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2446-2454. Tan, M., & Le, Q. V. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. Proceedings of the International Conference on Machine Learning (ICML), 6105-6114. Welch, G., & Bishop, G. (2006a). An Introduction to the Kalman Filter. UNC-Chapel Hill, TR 95-041. Welch, G., & Bishop, G. (2006b). An Introduction to the Kalman Filter (inf. téc. N.º TR 95-041). University of North Carolina at Chapel Hill. Wojke, N., Bewley, A., & Paulus, D. (2017). Simple Online and Realtime Tracking with a Deep Association Metric. IEEE International Conference on Image Processing (ICIP), 3645-3649. https://doi.org/10.1109/ICIP.2017.8296962 World Health Organization. (2023). Global Status Report on Road Safety 2023. Zhang, Y., Sun, P., Jiang, Y., Yu, D., Weng, F., Yuan, Z., Luo, P., Liu, W., & Wang, X. (2022). ByteTrack: Multi-Object Tracking by Associating Every Detection Box. European Conference on Computer Vision (ECCV), 1-21. https://doi.org/10.1007/978-3-031-20047-2_1 Zhang, Y., Wang, C., Wang, X., Zeng, W., & Liu, W. (2024). A Survey on Multi-Object Tracking: Methods, Datasets, and Metrics. ACM Computing Surveys, 56(3), 1-36. https://doi.org/10.1145/3630104 Zheng, L., Shen, L., Tian, L., Wang, S., Wang, J., & Tian, Q. (2015). Scalable Person Re-identification: A Benchmark. Proceedings of the IEEE International Conference on Computer Vision (ICCV), 1116-1124. Zhou, K., Yang, Y., Cavallaro, A., & Xiang, T. (2019). Omni-Scale Feature Learning for Person Re-Identification. IEEE/CVF International Conference on Computer Vision (ICCV), 3702-3712. | |
| dc.rights.accessrights | info:eu-repo/semantics/openAccess | |
| dc.rights.coar | http://purl.org/coar/access_right/c_abf2 | |
| dc.rights.creativecommons | Atribución-NoComercial-SinDerivadas 4.0 Internacional (CC BY-NC-ND 4.0) | |
| dc.rights.license | Atribución-NoComercial-CompartirIgual 4.0 Internacional (CC BY-NC-SA 4.0) | |
| dc.rights.local | Abierto (Texto Completo) | |
| dc.rights.uri | https://creativecommons.org/licenses/by-nc-sa/4.0/ | |
| dc.subject.armarc | Accidentes de tránsito | spa |
| dc.subject.armarc | Seguridad vial | spa |
| dc.subject.armarc | Innovaciones tecnológicas | spa |
| dc.subject.lemb | Mejoramiento de procesos | spa |
| dc.title | Uso de videoanalı́tica y modelos matemáticos para la detección y seguimiento de objetos para apoyar la gestión vial en la ciudad de Bogotá | spa |
| dc.title | Use of video analytics and mathematical models for object detection and tracking to support road management in the city of Bogotá | eng |
| dc.type | Trabajo de grado - Maestría | spa |
| dc.type.coar | http://purl.org/coar/resource_type/c_bdcc | |
| dc.type.coarversion | http://purl.org/coar/version/c_ab4af688f83e57aa | |
| dc.type.content | Text | |
| dc.type.driver | info:eu-repo/semantics/masterThesis | |
| dc.type.other | Trabajo de grado - Maestría | |
| dc.type.redcol | http://purl.org/redcol/resource_type/TM | |
| dc.type.version | info:eu-repo/semantics/acceptedVersion | |
| dspace.entity.type | Publication | |
| person.affiliation.name | Maestría en Ciencias de Datos | |
| relation.isDirectorOfPublication | 236dc412-a9d7-4a82-8979-861a50ea3fc6 | |
| relation.isDirectorOfPublication | 236dc412-a9d7-4a82-8979-861a50ea3fc6 | |
| relation.isDirectorOfPublication.latestForDiscovery | 236dc412-a9d7-4a82-8979-861a50ea3fc6 |
Archivos
Bloque de licencias
1 - 1 de 1
Cargando...
- Nombre:
- license.txt
- Tamaño:
- 1.92 KB
- Formato:
- Item-specific license agreed upon to submission
- Descripción:
