Tunnel Accident Warning Sign Detection Dataset

In the current transportation industry, tunnel accidents occur frequently, posing a great challenge to road safety. Although there are some monitoring systems, most cannot detect the status of warning signs in real time, resulting in slow responses. This dataset aims to enhance the recognition ability of object detection algorithms through high-quality annotated data, solving the problem of inaccurate recognition of warning signs in tunnels. Data is collected using high-resolution cameras under various lighting and weather conditions to ensure data diversity. Quality control measures include multiple rounds of annotation and expert reviews to ensure annotation accuracy and consistency. Data is stored in JPG format, accompanied by JSON files that record annotation information, facilitating subsequent processing and analysis. The core advantage of this dataset is its high data quality, with annotation accuracy of over 95% and good consistency. The use of new data augmentation techniques improves the generalization ability of the model. Experiments show that compared to traditional datasets, the trained model's accuracy in tunnel sign recognition tasks increased by 15%. Additionally, the application value of this dataset lies in significantly reducing traffic accidents and improving traffic safety.