Embedded ResNet-Based Semantic Segmentation in a robotic vehicle for Search and Rescue
DOI:
https://doi.org/10.17979/ja-cea.2026.47.13810Keywords:
perception and sensing, Robotics technology, field robotics, Autonomous Mobile Robots, Machine learning, Autonomous robotic systemsAbstract
Visual perception in search-and-rescue robotics must interpret unstructured scenes under varying lighting conditions while maintaining real-time performance on embedded hardware. This work presents the deployment of SARNet, a semantic segmentation network designed for disaster-response scenarios, on an ARGO J8 unmanned ground vehicle using an attention-based architecture. To achieve this, the network was adapted through retraining with oversampling techniques to balance underrepresented critical classes, and the model was integrated into a ROS 2 framework running on an NVIDIA Jetson AGX Orin. The system acquires RGB images from a ZED 2i camera, generates multi-class semantic masks, estimates the distance and orientation of detected individuals, and displays the information through a visual interface to assist the operator. The complete pipeline operates at 16.4 FPS, although it can exceed 30 FPS when the limitations imposed by the input sensor and post-processing stages are removed. The source code is available at: https://github.com/lucasjuradomartinez-cyber/TFGLucas_ws
References
Alonso, I., Riazuelo, L., Murillo, A.C., 2020. MiniNet: an efficient semantic segmentation ConvNet for real-time robotic applications. IEEE Transactions on Robotics 36, 1340 – 1347. doi:10.1109/TRO.2020.2974099.
Bravo-Arrabal, J., Vázquez-Martín, R., Fernández-Lozano, J.J., García-Cerezo, A., 2026. Strengthening multi-robot systems for search and rescue: Co-designing robotics and communications toward 6g. IEEE Communications Standards Magazine 10, 305–312. doi:10.1109/MCOMSTD.2026.3678471.
Buda, M., Maki, A., Mazurowski, M.A., 2018. A systematic study of the class imbalance problem in convolutional neural networks. Neural Networks 106, 249–259. doi:10.1016/j.neunet.2018.07.011.
Casper, J., Murphy, R.R., 2003. Human-robot interactions during the robot-assisted urban search and rescue response at the world trade center. IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics 33, 367–385. doi:10.1109/TSMCB.2003.811794.
De Nóbrega-Buyón, K., Fernández-Lozano, J.J., Vázquez-Martín, R., 2025. Modelado y simulación de un robot móvil en entornos naturales mediante ROS2 y Unity3D. Jornadas de Automática 46. doi:10.17979/ja-cea.2025.46.12187.
Guo, X., Liu, T., Li, Y., Lin, Z., Deng, Z., 2025. TUNI: Real-time rgb-t semantic segmentation with unified multi-modal feature extraction and cross-modal feature fusion. arXiv preprint arXiv:2509.10005 doi:10.48550/arXiv.2509.10005.
Hartley, R., Zisserman, A., 2004. Multiple View Geometry in Computer Vision. 2 ed., Cambridge University Press.
He, K., Zhang, X., Ren, S., Sun, J., 2016. Deep residual learning for image recognition, in: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770–778. doi:10.1109/CVPR.2016.90.
Morales, J., Vázquez-Martín, R., Mandow, A., Morilla-Cabello, D., García-Cerezo, A., 2021. The UMA-SAR Dataset: Multimodal data collection from a ground vehicle during outdoor disaster response training exercises. The International Journal of Robotics Research 40, 835–847. doi:10.1177/02783649211004959.
Murphy, R.R., 2014. Disaster Robotics. The MIT Press. doi:10.7551/mitpress/9407.001.0001.
Orbea, D., Ulloa, C.C., Cerro, J.d., Barrientos, A., 2024. Mobile victim signs monitoring through non-invasive robotic system. Lecture Notes in Networks and Systems 1114 LNNS, 141 – 153. doi:10.1007/978-3-031-70722-3_15.
Ramon-Soria, P., Gomez-Tamm, A., Garcia-Rubiales, F., Arrue, B., Ollero, A., 2019. Autonomous landing on pipes using soft gripper for inspection and maintenance in outdoor environments, in: IEEE International Conference on Intelligent Robots and Systems, pp. 5832 – 5839. doi:10.1109/IROS40897.2019.8967850.
Ronneberger, O., Fischer, P., Brox, T., 2015. U-net: Convolutional networks for biomedical image segmentation, in: Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, Springer. pp. 234–241. doi:10.1007/978-3-319-24574-4_28.
Salas-Espinales, A., Vázquez-Martín, R., García-Cerezo, A., Mandow, A., 2023. SAR Nets: An Evaluation of Semantic Segmentation Networks with Attention Mechanisms for Search and Rescue Scenes, in: IEEE International Symposium on Safety, Security, and Rescue Robotics, IEEE. pp. 139–144. doi:10.1109/SSRR56537.2023.10400741.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Lucas Jurado-Martínez, Ricardo Vázquez Martín, Anthony Mandow

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.