FIRESENSE: Video Database for Flame and Smoke Detection Algorithms

The FIRESENSE database contains 49 videos for testing fire detection algorithms, developed within the FP7-ENV-244088 FIRESENSE project. It includes 11 positive and 16 negative videos for flame detection, and 13 positive and 9 negative videos for smoke detection. The database was created by Nikos Grammalidis of the Information Technologies Institute and results using it are presented in published research papers.

Use Cases

• Benchmarking flame detection algorithms based on the provided 11 positive and 16 negative video samples.
• Training and validating smoke detection models based on the 13 positive and 9 negative video samples.
• Developing spatio-temporal fire modeling techniques based on the dynamic texture of video sequences.
• Testing multi-sensor network concepts for environmental monitoring in cultural heritage areas.

Strengths

• Contains a curated set of 49 videos specifically for fire detection tasks.
• Provides a clear split of 27 videos for flame detection and 22 videos for smoke detection.
• Has been used for research published in peer-reviewed journals like IEEE TCSVT.

Limitations

• Row count and video duration are unknown, which may limit suitability assessment.
• Column-level documentation is absent; video metadata and annotation details must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Information Technologies Institute, developed within the FP7-ENV-244088 FIRESENSE project.

Collection Method

Created for the 'Fire Detection and Management through a Multi-Sensor Network for the Protection of Cultural Heritage Areas' project.