A Novel Defocus-Degree-Based Phase Unwrapping and Fusion Algorithm for High-Speed and Large-Depth-Range 3D Measurement

Abstract

Binary defocusing technique is typically used in high-speed 3-D measurement. As the defocus degree varies with the measuring depth range, current defocusing technique has limitations while measuring large-depth-range (LDR) scene. Addressing this difficulty, this article proposes a novel defocus-degree-based phase unwrapping and phase fusion algorithm for LDR 3-D measurement. Firstly, we establish a mathematical model to present the relationship between defocus degree and phase, and present the defocus-degree-based phase unwrapping algorithm to directly obtain its accurate fringe order. Then, we derive a phase error model to estimate its phase error, and develop a phase fusion algorithm to obtain the lowest phase error by fusing two frequency phases based on the estimated error. With implementation of the proposed phase unwrapping and fusion algorithm, experiments are performed and the results demonstrate that our method can achieve a high-speed, high-precision and low-outlier 3-D measurement for LDR scene.