Model-Free Lens Distortion Correction Based on Phase Analysis of Fringe-Patterns

Area of Science:

• Computer Vision
• Image Processing
• Optical Engineering

Background:

• Existing lens correction methods rely on specific distortion models, which can fail if an unsuitable model is chosen.
• Model-based approaches have inherent limitations and potential drawbacks due to model selection inaccuracies.
• Accurate lens distortion correction is crucial for various applications, including photogrammetry and machine vision.

Purpose of the Study:

• To propose and validate a model-free lens distortion correction method.
• To overcome the limitations of traditional model-dependent distortion correction techniques.
• To accurately determine and correct lens distortion without prior assumptions about the distortion model.

Main Methods:

• Developed a theoretical framework establishing the relationship between distortion displacement and the phase of sinusoidal fringe-patterns.
• Employed phase demodulation analysis of fringe-patterns to generate a point-by-point distortion displacement map.
• Utilized instantaneous frequency distribution to automatically determine the distortion center for optimal correction.

Main Results:

• Successfully demonstrated a model-free approach to image correction based on the derived distortion displacement map.
• Validated the proposed method through numerical simulations and experimental tests using a wide-angle lens.
• Achieved accurate distortion correction without the need for predefined lens distortion models.

Conclusions:

• The proposed model-free method based on fringe-pattern phase analysis offers a robust solution for lens distortion correction.
• This technique effectively overcomes the drawbacks associated with model-dependent correction methods.
• The automatic determination of the distortion center further enhances the accuracy and applicability of the method.