Subway Guide Sign Classification Dataset

The current transportation industry faces challenges such as sign diversity and environmental complexity in the classification and recognition of subway guide signs. Existing solutions often rely on manual labeling, which is inefficient and prone to errors. This dataset aims to provide a high-quality, rich image library of subway guide signs to support the research and development of automated recognition technologies. Data collection was conducted using professional cameras in actual subway environments to ensure diversity in real scenes. To ensure data quality, multiple rounds of labeling and consistency checks were implemented, and all data were reviewed by experts. Data is stored in JPG format and organized by category. The core advantage of this dataset is its high labeling accuracy and consistency, with labeling accuracy exceeding 95%. In addition, advanced image enhancement techniques have been used to improve the model's generalization ability, with an expected improvement of at least 20% in guide sign recognition performance metrics.