Underwater 3D reconstruction using a photometric stereo with illuminance estimation.

Area of Science:

• Optics and Photonics
• Computer Vision
• Robotics

Background:

• Underwater image sensing is challenged by light refraction and changing illuminance.
• Accurate 3D reconstruction is crucial for underwater tasks like defect detection.

Purpose of the Study:

• To analyze near-field light source convergence in water using a light propagation model.
• To develop an accurate photometric stereo (PS) method for underwater 3D reconstruction.
• To improve surface microdefect detection in underwater objects.

Main Methods:

• Utilized a light propagation model to analyze near-field point light source convergence in water.
• Determined the photometric stereo (PS) formula based on estimated camera illuminance.
• Designed and implemented an underwater PS system for experimental validation.

Main Results:

• Demonstrated improved accuracy in normal vector calculations.
• Validated the feasibility of the proposed underwater PS system.
• Achieved more precise underwater 3D reconstruction of objects.

Conclusions:

• The developed method enhances the accuracy of underwater 3D reconstruction.
• The system is suitable for detecting surface microdefects on underwater objects.
• Accurate illuminance estimation is key to improving underwater optical sensing.