Dynamic Monitoring Dataset of Open Flame Spread

The current industrial sector faces significant challenges in fire safety management, especially in environments such as production areas, cable trenches, and storage rooms. Early detection of open flames is crucial to preventing fire spread. Existing fire monitoring systems often rely on static surveillance and cannot effectively capture dynamic changes. This dataset aims to solve the technical difficulties in real-time monitoring by providing a wealth of dynamic video cases to meet companies' urgent needs for safe production and fire warning. Data collection is conducted using high-resolution cameras in various environments. To ensure data quality, multiple rounds of annotation and expert reviews are conducted to ensure the accuracy and consistency of the video data. Data is stored in MP4 format, with structured storage of video ID, timestamp, event tags, and location information, facilitating subsequent analysis and model training.