Automatic water-level measurement system for confined-space applications.

Area of Science:

• Environmental monitoring
• Computer vision
• Hydrology

Background:

• Flooding poses significant risks to life and property.
• Surveillance cameras are crucial for monitoring river water levels.
• Existing methods may suffer from inaccuracies due to camera angles.

Purpose of the Study:

• To develop an automatic water-level measurement system using a single camera.
• To address and correct measurement errors caused by perspective distortion.
• To enhance the accuracy of river water level monitoring.

Main Methods:

• Utilizing basic digital image processing to identify water marks on staff gauges.
• Employing image histograms for water level determination.
• Implementing inverse perspective mapping to rectify perspective distortion.

Main Results:

• The system accurately identifies preliminary water mark positions.
• Image histograms provide a reliable reference for water level calculation.
• Inverse perspective mapping effectively reduces measurement errors from perspective distortion.

Conclusions:

• The proposed system accurately measures river water levels in confined spaces.
• It significantly reduces image measurement errors caused by perspective distortion.
• This technology offers a reliable solution for automated flood prevention monitoring.