A two-stage context-aware meta-ensemble framework for collision risk prediction in Internet of Vehicles

Abstract

A key challenge in Internet of Vehicles (IoV) environments is accurately predicting collision risks, given the dynamic nature of traffic systems and the variety of available data sources. Existing approaches often rely on structured IoV data, while visual contextual information is not fully incorporated into the prediction process. This limitation motivates a unified framework that integrates environmental perception with sensor-derived information to enable more accurate collision-risk prediction. In this study, we propose a two-stage context-aware meta-ensemble framework that combines image-based perception and IoV-based predictive analysis for collision-risk prediction. The first stage uses multiple pre-trained convolutional neural networks to obtain rich visual representations from vehicle images through feature-level fusion. The second stage employs a hybrid stacked ensemble architecture that combines the probabilistic outputs of heterogeneous machine learning models with IoV and image-derived features, which a meta-learner then processes for final collision-risk prediction. The proposed feature-fusion model performs best in the perception stage, with an accuracy of 0.97. In the IoV-based prediction stage, the proposed stacked ensemble architecture outperforms all baseline models, achieving an accuracy of 0.9847. Robustness analysis shows that the perception module tolerates blur and rotation but remains sensitive to Gaussian noise. Cross-stage integration incorporates image-derived features into the IoV feature space, providing complementary visual context for collision-risk prediction and demonstrating its potential for intelligent transportation and cyber-physical IoV applications.

Keywords

Internet of Vehicles / collision risk prediction / cyber-physical systems / feature-level fusion / deep learning / explainable artificial intelligence / robustness analysis

References

[1] Biswas A.Autonomous vehicles enabled by the integration of IoT, edge intelligence, 5G, and blockchain.Sensors2023;23:1963 PMCID:PMC9963447

[2] Kumar N,Saini M.IoV-fog-assisted framework for accident detection and classification.ACM Trans Embed Comput Syst2025;24:1-19

[3] Mohialden YM,Hussain SAK.Mitigating cybersecurity threats in autonomous vehicles: a real-world case study on protecting V2X communication against remote hacking attempts.Cybersys J2024;1:41-50

[4] Rishiwal V,Alotaibi A,Yadav P.Exploring secure V2X communication networks for human-centric security and privacy in smart cities.IEEE Access2024;12:138763-88

[5] Ouaissa M,Houmer M,Boulouard Z.A secure vehicle to everything (V2X) communication model for intelligent transportation system. In: Ouaissa, M.; Boulouard, Z.; Ouaissa, M.; Guermah, B.; eds. Computational intelligence in recent communication networks. Cham: Springer International Publishing; 2022. pp. 83-102.

[6] Alharthi M,Djenouri D.Ensemble learning approaches for multi-class intrusion detection systems for the Internet of Vehicles (IoV): a comprehensive survey.Future Internet2025;17:317

[7] Seo JK,Kim B,Seo JT.AI-based anomaly detection in industrial control and cyber–physical systems: a data-type-oriented systematic review.Electronics2026;15:20

[8] Alsulami AA,Alturki B,Alsini R.Security strategy for autonomous vehicle cyber-physical systems using transfer learning.J Cloud Comp2023;12:564

[9] Xing W.Security control of cyber-physical systems under cyber attacks: a survey.Sensors2024;24:3815 PMCID:PMC11207848

[10] Mienye ID.A comprehensive review of deep learning: architectures, recent advances, and applications.Information2024;15:755

[11] Salim MM,Loia V.Deep learning enabled secure IoT handover authentication for blockchain networks.Hum Cent Comput Inf Sci2021;11:21

[12] Mozaffaree Pour, N.; Oja, T. Prediction power of logistic regression (LR) and multi-layer perceptron (MLP) models in exploring driving forces of urban expansion to be sustainable in Estonia.Sustainability2022;14:160

[13] Shah K,Sanghvi D.A comparative analysis of logistic regression, random forest and KNN models for the text classification.Augment Hum Res2020;5:32

[14] Algarni A,Alaa Ala’anzy M.An edge computing-based and threat behavior-aware smart prioritization framework for cybersecurity intrusion detection and prevention of IEDs in smart grids with integration of modified LGBM and one class-SVM models.IEEE Access2024;12:104948-63

[15] Safi A,Jehangiri AI.A fault tolerant surveillance system for fire detection and prevention using LoRaWAN in smart buildings.Sensors2022;22:8411 PMCID:PMC9657799

[16] Talaat FM.Effective prediction and resource allocation method (EPRAM) in fog computing environment for smart healthcare system.Multimed Tools Appl2022;81:8235-58

[17] Muthuramalingam S,Rakesh Kumar S,Sathiyaraj R.IoT based intelligent transportation system (IoT-ITS) for global perspective: a case study. In: Balas, V. E.; Solanki, V. K.; Kumar, R.; Khari, M.; eds. Internet of Things and big data analytics for smart generation. Cham: Springer International Publishing; 2019. pp. 279-300.

[18] Balajee A,Suresh A,Mahesh T.Federated hyper LSTM model for storage optimization and collision prediction in an intelligent IoVT.Egypt Inform J2026;33:100884

[19] Seghir Z,Barka K,Toral-Cruz H.A real-time consensus-free accident detection framework for Internet of Vehicles using vision transformer and EfficientNet.AI2026;7:4

[20] Li H.Traffic accident risk prediction based on deep learning and spatiotemporal features of vehicle trajectories.PLoS One2025;20:e0320656 PMCID:PMC12048164

[21] Khan FM,Rahman T.Federated deep learning for collision avoidance in IoV with digital twin integration.Expert Syst2026;43:e70168

[22] Selvaraj DC,Panayiotou T,Chiasserini CF.Edge-assisted ML-aided uncertainty-aware vehicle collision avoidance at urban intersections.IEEE Trans Intell Veh2026;11:573-89

[23] Liu Y,Lin P,Li K.A framework for real-time vehicle tracking in large-scale roadside sensor networks.Green Energy Intell Transp2025;4:100362

[24] Zhang J,Wang Y.Optical communication based V2V for vehicle platooning.Green Energy Intell Transp2025;4:100278

[25] Khan MAA,Farhad A,Mirani IK.When LoRa meets distributed machine learning to optimize the network connectivity for green and intelligent transportation system.Green Energy Intell Transp2024;3:100204

[26] Kaggle. Vehicle Images Dataset. https://www.kaggle.com/datasets/lyensoetanto/vehicle-images-dataset. (accessed 2026-09-15)

[27] Kaggle. Internet of Vehicles Dataset. https://www.kaggle.com/datasets/harunachiromagombe/internet-of-vehicles-dataset. (accessed 2026-09-15)