Determining scale and sea state from water video.

Area of Science:

• Computer Vision
• Oceanography
• Image Processing

Background:

• Determining real-world scale in image processing typically requires camera calibration.
• Dynamic scenes, particularly ocean environments, present unique challenges for scale estimation.
• Existing methods often fail in uncalibrated or dynamic scenarios.

Purpose of the Study:

• To develop a method for recovering real-world scale and sea state from uncalibrated camera footage of ocean scenes.
• To utilize the inherent physics of water waves for accurate environmental measurements.
• To enable applications in maritime surveillance and monitoring without prior calibration.

Main Methods:

• Applying Fourier transforms to spatial and temporal dimensions to obtain frequency spectra.
• Utilizing the water wave dispersion relation (period-wavelength relationship) to infer scale.
• Integrating stochastic oceanography models with spectral analysis to determine sea state parameters.

Main Results:

• Successfully recovered real-world scale and sea state parameters (peak wavelength, period, wave height, estimated wind speed) from ocean sequences.
• Validated the method on both synthetic and real-world video data against known scene geometry.
• Demonstrated the effectiveness of the dispersion relation in uncalibrated scale recovery.

Conclusions:

• The proposed method accurately estimates real-world scale and sea state in dynamic ocean scenes using uncalibrated cameras.
• This technique offers a viable solution for applications requiring environmental measurements in maritime settings.
• The findings have significant implications for port monitoring, coastal surveillance, and marine research.