Vehicle Damage Detection After Accidents Using Deep Neural Networks: A Case Study in the Insurance Industry
Subject Areas :
Fatemeh sarami
1
,
Hossein Mohammadi Dolat Abadi
2
*
1 - Industrial engineering, College of Farabi, university of tehran
2 - university of tehran
Keywords: Deep Learning, Image Processing, Convolutional Neural Network, Vehicle Damage, Insurance Industry,
Abstract :
Objective: This study aims to investigate the capability of convolutional neural networks (CNNs) to identify and classify vehicle damage after accidents and to evaluate their potential application in the insurance industry. Necessity: Vehicle damage assessment in the insurance industry is generally time-consuming and requires expert inspection. The application of intelligent image-processing methods can facilitate preliminary damage assessment, reduce claim-processing time, and support the development of intelligent insurance systems. Methodology: The research data consists of real images of damaged vehicles belonging to third-party insurance customers during 2020 and 2021. From approximately 20,000 images, 4,100 were selected to train and evaluate seven deep learning architectures: AlexNet, VGG-19, ResNet-50, ResNet-101, EfficientNetB7, EfficientNetV2L, and MobileNetV2. The models were evaluated for classifying six types of vehicle damage: superficial damage, severe damage, side-mirror damage, windshield damage, tire damage, and vehicle-light damage. Findings: The results showed that model performance in terms of speed and accuracy was influenced by the structure and characteristics of the pre-trained architecture. The classification accuracy of the evaluated models ranged from 59% to 63%. The findings also demonstrated the capability of deep convolutional neural networks to extract visual features and distinguish different types of vehicle damage. Conclusion: The findings indicate that convolutional neural networks and transfer learning methods have promising potential for the identification and preliminary assessment of vehicle damage after accidents. These models can provide a foundation for developing intelligent damage assessment systems in the insurance industry, helping to reduce assessment time, accelerate claim-processing procedures, and improve the efficiency of insurance processes
Albelwi, S., & Mahmood, A. (2017). A framework for designing the architectures of deep convolutional neural networks. Entropy, 19(6), 242.
Amin, S., Hijji, M., Iqbal, R., Harrop, W., & Chang, V. (2019). Fuzzy expert system-based framework for flood management in Saudi Arabia. Cluster Computing, 22, 11723–11740.
Bhandare, A., Bhide, M., Gokhale, P., & Chandavarkar, R. (2016). Applications of convolutional neural networks. International Journal of Computer Science and Information Technologies, 7(5), 2206–2215.
Derrig, R. A. (2002). Insurance fraud. Journal of Risk and Insurance, 69(3), 271–287.
Dorafshan, S., Thomas, R. J., & Maguire, M. (2018). Comparison of deep convolutional neural networks and edge detectors for image-based crack detection in concrete. Construction and Building Materials, 186, 1031–1045.
Eom, H., & Choi, H. (2018). Alpha-pooling for convolutional neural networks. arXiv Preprint arXiv:1811.03436.
Gao, Y., & Mosalam, K. M. (2018). Deep transfer learning for image-based structural damage recognition. Computer-Aided Civil and Infrastructure Engineering, 33(9), 748–768.
Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.
Gopalakrishnan, K. (2018). Deep learning in data-driven pavement image analysis and automated distress detection: A review. Data, 3(3), 28.
Gopalakrishnan, K., Gholami, H., Vidyadharan, A., Choudhary, A., & Agrawal, A. (2018). Crack damage detection in unmanned aerial vehicle images of civil infrastructure using pre-trained deep learning model. International Journal of Traffic and Transportation Engineering, 8(1), 1–14.
Gopalakrishnan, K., Khaitan, S. K., Choudhary, A., & Agrawal, A. (2017). Deep convolutional neural networks with transfer learning for computer vision-based data-driven pavement distress detection. Construction and Building Materials, 157, 322–330.
Han, J., Pei, J., & Tong, H. (2022). Data Mining: Concepts and Techniques. Morgan Kaufmann.
Harshani, W. R., & Vidanage, K. (2017). Image processing based severity and cost prediction of damages in the vehicle body: A computational intelligence approach. In 2017 National Information Technology Conference (NITC). IEEE.
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 770–778).
Kim, H., Kim, H., Hong, Y. W., & Byun, H. (2018). Detecting construction equipment using a region-based fully convolutional network and transfer learning. Journal of Computing in Civil Engineering, 32(2), 04017082.
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems, 25, 1097–1105.
Kumar, N., Srivastava, J. D., & Bisht, H. (2019). Artificial intelligence in insurance sector. Journal of the Gujarat Research Society, 21(7), 79–91.
LeCun, Y., Bengio, Y., & Hinton, G. (2015). Deep learning. Nature, 521(7553), 436–444.
Li, B., Wang, K. C., Zhang, A., Yang, E., & Wang, G. (2020). Automatic classification of pavement crack using deep convolutional neural network. International Journal of Pavement Engineering, 21(4), 457–463.
Li, L., Jamieson, K., DeSalvo, G., Rostamizadeh, A., & Talwalkar, A. (2017). Hyperband: A novel bandit-based approach to hyperparameter optimization. Journal of Machine Learning Research, 18(1), 6765–6816.
Li, P., Shen, B., & Dong, W. (2018). An anti-fraud system for car insurance claim based on visual evidence. arXiv Preprint arXiv:1804.11207.
Maiano, L., Montuschi, A., Caserio, M., Ferri, E., Kieffer, F., Germanò, C., Baiocco, L., Celsi, L. R., Amerini, I., & Anagnostopoulos, A. (2023). A deep-learning-based antifraud system for car-insurance claims. Expert Systems with Applications, 120644.
Najafabadi, M. M., Villanustre, F., Khoshgoftaar, T. M., Seliya, N., Wald, R., & Muharemagic, E. (2015). Deep learning applications and challenges in big data analytics.