Multi-Frequency Nonlinear Methods for 3D Shape Measurement of Semi-Transparent Surfaces Using Projector-Camera...

Area of Science:

• Computer Vision
• Metrology
• Optical Engineering

Background:

• 3D shape measurement of semi-transparent surfaces is challenging due to diffuse light reflection and transmission.
• Participating background surfaces further complicate accurate 3D reconstruction.

Purpose of the Study:

• To develop and evaluate novel methods for precise 3D shape measurement of semi-transparent surfaces.
• To address limitations of existing 3D scanning techniques in challenging optical conditions.

Main Methods:

• Two methods using sinusoidal patterns with frequencies optimized for projector-camera systems.
• Method 1: Discrete Fourier Transform (DFT) on pixel intensity signals to map camera pixels to projector columns.
• Method 2: Analytical model fitting to pixel intensity signals for camera-projector correspondence, followed by surface continuity constraints and triangulation.

Main Results:

• Both methods successfully establish camera-projector correspondence, enabling 3D point cloud generation.
• Experimental results validate the accuracy and reliability of the proposed 3D scanning techniques.
• The methods effectively handle semi-transparent surfaces and complex backgrounds.

Conclusions:

• The proposed DFT and analytical model-based methods offer robust solutions for 3D shape measurement of semi-transparent objects.
• These advancements enhance the performance and applicability of projector-camera 3D scanning systems.
• Accurate 3D reconstruction is achieved through effective handling of optical challenges.