Water Meter Images with OCR Labels and Segmentation Masks

Over 5,000 images of water meters, including segmentation masks and OCR labels for meter readings. The dataset is designed for research in water consumption analysis and smart meter technology, providing insights into residential and commercial water usage. It was created by UniDataPro and last updated on 2025-04-04.

Use Cases

• Train segmentation models to identify water meter components based on the provided masks.
• Develop OCR systems to automatically extract numeric readings from water meter images.
• Analyze patterns in residential and commercial water consumption using the labeled meter readings.

Strengths

• Contains over 5,000 images, providing a substantial base for model training.
• Includes corresponding segmentation masks and OCR labels, offering multi-task learning opportunities.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

UniDataPro on Hugging Face

Collection Method

Likely collected and annotated for computer vision research.

Freshness

Last updated 2025-04-04 11:32:07; freshness should be verified.