Image-Based Crash Severity Analysis: Transforming Tabular Data Using Convolutional Neural Networks
Open Access
- 23 February 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 33665-33683
Abstract
Traffic safety is a critical issue in urban transportation systems, with crash severity predictions playing a key role in mitigating road traffic injuries and fatalities. This study addresses the research question: how can transforming tabular crash data into image-based representations using deep learning techniques improve the prediction of crash severity, while effectively managing class imbalances? For analysis, 63,745 barrier crashes that occurred in Texas from 2017–2022 were examined. Using a sample database of barrier crash incidents, we introduce a novel approach that leverages Convolutional Neural Networks (CNNs) to convert traditional tabular crash data into image representations by employing the DeepInsight technique, which maps feature similarity into structured two-dimensional grids, enhancing the accuracy and effectiveness of prediction models. Generative Adversarial Networks (GANs) were used for data augmentation, addressing issues of class imbalances with innovative sampling techniques including under-sampling, over-sampling, and balanced sampling. Key findings indicate that balanced datasets significantly improve predictive performance, enhancing generalizability across diverse crash scenarios. Moreover, under- and over-sampling strategies offer a refined trade-off among precision, recall, and F1-score, while the over-sampling method distinctly excels in practical model performance by reducing misclassification rates and improving overall accuracy. Comparative evaluation against an XGBoost baseline on identical balanced data highlights the ability of the image-based deep learning framework to capture complex crash patterns. These findings suggest that integrating feature-similarity–based tabular-to-image transformation with deep learning and balanced data sampling can significantly improve crash severity classification.Keywords
This publication has 54 references indexed in Scilit:
- Comparison Performance of Spectrogram and Scalogram as Input of Acoustic Recognition Task2025
- A new spatiotemporal convolutional neural network model for short-term crash predictionFrontiers of Engineering Management, 2024
- A Deep Learning Framework for Generation and Analysis of Driving Scenario TrajectoriesSN Computer Science, 2023
- Image-Based Sentiment Analysis Using InceptionV3 Transfer Learning ApproachSN Computer Science, 2023
- Highway crash detection and risk estimation using deep learningAccident Analysis & Prevention, 2020
- Real-time crash risk prediction on arterials based on LSTM-CNNAccident Analysis & Prevention, 2020
- Barrier-relevant crash modification factors and average costs of crashes on arterial roads in IndianaAccident Analysis & Prevention, 2018
- An insight into the performance of road barriers − redistribution of barrier-relevant crashesAccident Analysis & Prevention, 2016
- Median barrier crash severity: Some new insightsAccident Analysis & Prevention, 2010
- Predicting the frequency of median barrier crashes on Pennsylvania interstate highwaysAccident Analysis & Prevention, 2006