Developing an Optical Image-Based Method for Bridge Deformation Measurement Considering Camera Motion

Area of Science:

• Structural Engineering
• Computer Vision
• Photogrammetry

Background:

• Camera vibrations during image acquisition can introduce errors in bridge deformation estimation.
• Accurate measurement of bridge deformation is crucial for structural health monitoring and safety assessments.

Purpose of the Study:

• To develop and validate a methodology for removing image global motion to improve bridge deformation measurement accuracy.
• To integrate an image motion correction algorithm with a 2D image-based deformation measurement technique.

Main Methods:

• Utilized an Iterative Affine Motion Estimator on a defined sub-image to estimate motion parameters.
• Applied estimated parameters to correct global motion in all pixels of captured images.
• Employed a 2D image-based deformation measurement technique on corrected images to track targets and measure deformation.

Main Results:

• The proposed methodology successfully removed image global motion artifacts.
• Validated through laboratory and field experiments, demonstrating accuracy and reliability.
• Enabled precise extraction and measurement of real bridge deformation.

Conclusions:

• The combined approach of image motion correction and 2D deformation analysis is effective for accurate bridge monitoring.
• This technique enhances the reliability of deformation measurements in the presence of camera vibrations.
• The validated methodology offers a robust solution for structural health assessment of bridges.