Due to massive increase in road vehicles, the rate of traffic accidents increases rapidly. These accidents cause injuries, disabilities and may even result in loss of lives. To analyze traffic accident data, one needs to understand and formalize the dataset. Mostly, this data is stored in relational databases or in raw text format. This paper proposes a framework to store traffic accident data in a big data platform (Casandra). This will enable fast and accurate analyses technique that use big data platforms. The paper proposes applying deep learning techniques to build prediction and classification models from the road accident data. In particular, the models will focus on the prediction and identification of hotspots on the roads in order to decrease the number of accidents in the future.
Towards Deep Learning based Traffic Accident Analysis
2020-01-01
5890329 byte
Conference paper
Electronic Resource
English
Road traffic accident risk prediction deep learning algorithm
European Patent Office | 2022
|Traffic accident severity prediction based on interpretable deep learning model
Taylor & Francis Verlag | 2025
|Traffic accident prediction method and device based on deep learning
European Patent Office | 2024
|