Single-shot 3D motion picture camera with a dense point cloud

TL;DR

This paper presents a novel single-shot 3D camera that captures dense point clouds without pattern coding, achieving near-theoretical information efficiency and enabling 3D motion picture acquisition with high depth resolution.

Contribution

The paper introduces a pattern-free, single-shot 3D camera system utilizing triangulation and algorithms to resolve ambiguity, enabling dense 3D shape measurement and motion picture capture.

Findings

Captures up to 30% of camera pixels as independent 3D points.

Achieves depth resolution at physical noise limits.

Enables 3D motion-picture acquisition of fast processes.

Abstract

We introduce a method and a 3D-camera for single-shot 3D shape measurement, with unprecedented features: The 3D-camera does not rely on pattern codification and acquires object surfaces at the theoretical limit of the information efficiency: Up to 30% of the available camera pixels display independent (not interpolated) 3D points. The 3D-camera is based on triangulation with two properly positioned cameras and a projected multi-line pattern, in combination with algorithms that solve the ambiguity problem. The projected static line pattern enables 3D-acquisition of fast processes and the take of 3D-motion-pictures. The depth resolution is at its physical limit, defined by electronic noise and speckle noise. The requisite low cost technology is simple.