Learning-based reconstruction of moving transparent objects

This paper studies 3D reconstruction for moving transparent objects, a setting where refraction, weak surface texture, and motion make conventional structured-light reconstruction difficult.

This work appears in International Conference on Optoelectronic Science and Intelligent Sensing (SPIE), 1417511.

Algorithm principle

The method uses a learning-based reconstruction pipeline to compensate for the optical ambiguity of transparent objects and the temporal inconsistency caused by motion.

Visual material

Detailed method and result figures will be added when the original paper assets are available.

Results and impact

The work belongs to the structured-light and transparent-object reconstruction thread, where robustness matters for industrial surfaces that are glossy, transparent, or in motion.