Industrial Fan Vibration and Anomaly Recognition Dataset

In the current industrial sector, fans are critical equipment, and their operational stability directly affects production efficiency. However, fans often experience vibrations, abnormal sounds, and jumps during operation. If these phenomena are not identified and handled in time, they may lead to equipment damage and production stoppages. Existing monitoring solutions mainly rely on traditional sensor data, lacking comprehensive analysis of visual data, which leads to insufficient fault warning. This dataset aims to capture abnormal behaviors during fan operation through video data, enabling more precise fault identification and prediction. Data collection is conducted using high-frame-rate cameras in real industrial environments, ensuring the authenticity and diversity of the data. In terms of quality control, a multi-round annotation and expert review process is adopted to ensure consistency and accuracy of the annotations. The data is stored in MP4 format, organized by event type and timestamp, facilitating subsequent analysis and processing. The core advantage of this dataset is its high-quality annotation and rich video data, with annotation accuracy above 95%. By introducing advanced image processing and data augmentation techniques, it significantly enhances the model's generalization capabilities. Compared to traditional sensor data, this dataset improves the accuracy of anomaly detection by 20%.