Motion-oriented high speed 3-D measurements by binocular fringe projection using binary aperiodic patterns.

Area of Science:

• Optical Metrology
• 3-D Imaging
• Computer Vision

Background:

• High-speed 3-D measurement of dynamic objects is crucial in various fields.
• Traditional fringe projection techniques suffer from motion artifacts due to sequential pattern recording.
• Existing methods struggle to achieve pixel-level accuracy for moving subjects.

Purpose of the Study:

• To develop a novel high-speed 3-D scanning technique that overcomes motion-induced artifacts.
• To enhance the accuracy and robustness of 3-D reconstructions for dynamic objects.
• To achieve sub-pixel disparity results for precise 3-D measurements.

Main Methods:

• A novel high-speed 3-D scanning technique combining fringe projection and stereo vision.
• Modification of aperiodic sinusoidal patterns for kilohertz fringe projection using a defocusing strategy.
• Development of a temporal intensity tracing algorithm for motion influence mitigation.
• Implementation of a combined cost measure for robust pixel cost estimation.
• A three-step refinement framework for outlier elimination and sub-pixel disparity calculation.

Main Results:

• The proposed method achieves promising measuring speeds at kilohertz frequencies.
• Significant reduction in motion artifacts, leading to improved reconstruction accuracy.
• Experimental results demonstrate an order of magnitude improvement in accuracy for dynamic objects compared to traditional methods.
• Obtained sub-pixel disparity results for enhanced 3-D reconstruction quality.

Conclusions:

• The novel fringe projection and stereo-based technique effectively mitigates motion artifacts in high-speed 3-D scanning.
• The developed algorithms and framework enable accurate and robust 3-D measurements of dynamic objects.
• This advancement offers a significant improvement for applications requiring precise 3-D data of moving targets.