Design of diffractive optical element projector for a pseudorandom dot array by an improved encoding method

Miao, Yinxiao; Zhao, Yun; Ma, Hongyuan; Jiang, Minwei; Lin, Jie; Jin, Peng

doi:10.1364/ao.58.00g169

Cited by 10 publications

(5 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These three classic algorithms are the most representative method of its types, are widely cited and referenced. Other related methods are mostly based on these methods with minor changes to adapt to specific applications, refer to [2], [3], [12]- [14] et al…”

Section: Compared With Other Algorithmsmentioning

confidence: 99%

See 1 more Smart Citation

M-Array Based on Non-Zero Maps

et al. 2021

View full text Add to dashboard Cite

Two-dimensional maximum area array (m-array) is a typical pseudo-random array with specific window property -all sub-windows are globally unique. M-array constructed by Formula-Method-Based encoding algorithm features a big difference between the number of rows and columns. Enumeration-Method-Based encoding algorithm only constructs small-size m-array with high computational complexity and low success rate. Based on the characteristics of Galois Fields, a new method of m-array construction based on non-zero maps theory is proposed and the theoretical proof is presented. Given symbols p and sub-window size m × n as input, m-array with a large array size and close numbers of rows and columns is constructed by dividing and splicing m-sequence. Comparing with existing methods, the advantages of the algorithm include larger array size, more balanced number of rows and columns, and lower computational complexity. If needed, the method in this paper is flexible enough that the number of rows and columns of m-array can also be adjusted within a certain range, meaning that m-array with specific rows and columns ratio can be constructed.INDEX TERMS M-array, non-zero maps, pseudo-random array, window property.XIAO ZHOU received the Ph.D. degree in pattern recognition and intelligent system from Huazhong

show abstract

Section: Compared With Other Algorithmsmentioning

confidence: 99%

“…Based on the diagonal algorithm, Lu [2] selected a square sub-array as the main body and spliced the remaining sub-windows to m-array through matching. Miao [3] tried to optimize the diagonal algorithm by central symmetrizing the sub-array.…”

Section: Introductionmentioning

confidence: 99%

M-Array Based on Non-Zero Maps

et al. 2021

View full text Add to dashboard Cite

show abstract

“…However, this heuristic design is blind to scene statistics, noise level, and the reconstruction method. Existing methods have attempted to improve depth estimation by employing hand-crafted DOE designs [11,49,34] that rely on golden-eye experts and heuristic metrics on the illumination pattern, not directly correspond to system depth accuracy. We depart from these heuristic designs and instead directly optimize the illumination pattern with the final depth reconstruction accuracy as a loss via end-to-end optimization.…”

Section: Related Workmentioning

confidence: 99%

Polka Lines: Learning Structured Illumination and Reconstruction for Active Stereo

Baek¹,

Heide²

2020

Preprint

View full text Add to dashboard Cite

Active stereo cameras that recover depth from structured light captures have become a cornerstone sensor modality for 3D scene reconstruction and understanding tasks across application domains. Existing active stereo cameras project a pseudo-random dot pattern on object surfaces to extract disparity independently of object texture. Such hand-crafted patterns are designed in isolation from the scene statistics, ambient illumination conditions, and the reconstruction method. In this work, we propose the first method to jointly learn structured illumination and reconstruction, parameterized by a diffractive optical element and a neural network, in an end-to-end fashion. To this end, we introduce a novel differentiable image formation model for active stereo, relying on both wave and geometric optics, and a novel trinocular reconstruction network. The jointly optimized pattern, which we dub "Polka Lines," together with the reconstruction network, achieve state-of-the-art activestereo depth estimates across imaging conditions. We validate the proposed method in simulation and on a hardware prototype, and show that our method outperforms existing active stereo systems.

show abstract

“…[ 4–8 ] The baseline length refers to the distance between the projection device and the camera, which together with the optical element sizes, determines the overall volume of the sensing system. Recently, with the increasing demand for compact and portable optoelectronic systems in commercial and industrial fields, diffractive optical elements and metasurfaces have been widely adopted to project various SL images, [ 9–12 ] significantly reducing the size of projected light sources.…”

Section: Introductionmentioning

confidence: 99%

Metasurface‐Based Structured Light Sensing Without Triangulation

Li,

et al. 2023

Advanced Optical Materials

View full text Add to dashboard Cite

Structured light (SL) sensing technology, based on the triangulation principle, is widely employed for depth perception in various applications such as face recognition, human‐computer interaction, machine vision, and more. In this approach, the baseline length, which refers to the distance between the projection device and the camera, is a critical parameter that affects depth measurement and system miniaturization. Recently, metasurfaces have emerged as promising devices for constructing compact optoelectronic systems due to their excellent performance in wavefront modulation of light and ultra‐thin characteristics. In this study, a metasurface‐based monocular SL depth detection scheme is proposed that incorporates a specially designed 3D holographic projection. Under this projected light field, the entire 3D space is labeled, and depth information can be obtained without considering the baseline length, which can further reduce the volume of SL systems. Additionally, a template‐matching algorithm based on correlation coefficient analysis is developed and experimentally demonstrates its feasibility by precisely positioning objects. It is believed that this work opens up a new perspective for compact, lightweight, and flexible design of SL sensing systems, and has a promising future in quantitative detection, automatic location, and industrial measurement.

show abstract

Design of diffractive optical element projector for a pseudorandom dot array by an improved encoding method

Cited by 10 publications

References 27 publications

M-Array Based on Non-Zero Maps

M-Array Based on Non-Zero Maps

Polka Lines: Learning Structured Illumination and Reconstruction for Active Stereo

Metasurface‐Based Structured Light Sensing Without Triangulation

Contact Info

Product

Resources

About