Journals >Advanced Photonics Nexus
Phase unwrapping is an indispensable step for many optical imaging and metrology techniques. The rapid development of deep learning has brought ideas to phase unwrapping. In the past four years, various phase dataset generation methods and deep-learning-involved spatial phase unwrapping methods have emerged quickly. However, these methods were proposed and analyzed individually, using different strategies, neural networks, and datasets, and applied to different scenarios. It is thus necessary to do a detailed comparison of these deep-learning-involved methods and the traditional methods in the same context. We first divide the phase dataset generation methods into random matrix enlargement, Gauss matrix superposition, and Zernike polynomials superposition, and then divide the deep-learning-involved phase unwrapping methods into deep-learning-performed regression, deep-learning-performed wrap count, and deep-learning-assisted denoising. For the phase dataset generation methods, the richness of the datasets and the generalization capabilities of the trained networks are compared in detail. In addition, the deep-learning-involved methods are analyzed and compared with the traditional methods in ideal, noisy, discontinuous, and aliasing cases. Finally, we give suggestions on the best methods for different situations and propose the potential development direction for the dataset generation method, neural network structure, generalization ability enhancement, and neural network training strategy for the deep-learning-involved spatial phase unwrapping methods.
.Thin-film lithium niobate is a promising material platform for integrated nonlinear photonics, due to its high refractive index contrast with the excellent optical properties. However, the high refractive index contrast and correspondingly small mode field diameter limit the attainable coupling between the waveguide and fiber. In second harmonic generation processes, lack of efficient fiber-chip coupling schemes covering both the fundamental and second harmonic wavelengths has greatly limited the overall efficiency. We design and fabricate an ultra-broadband tri-layer edge coupler with a high coupling efficiency. The coupler allows efficient coupling of 1 dB / facet at 1550 nm and 3 dB / facet at 775 nm. This enables us to achieve an ultrahigh overall second harmonic generation normalized efficiency (fiber-to-fiber) of 1027 % W - 1 cm - 2 (on-chip second harmonic efficiency ∼3256 % W - 1 cm - 2) in a 5-mm-long periodically-poled lithium niobate waveguide, which is two to three orders of magnitude higher than that in state-of-the-art devices.
.Since Allen et al. demonstrated 30 years ago that beams with helical wavefronts carry orbital angular momentum (OAM), the OAM of beams has attracted extensive attention and stimulated lots of applications in both classical and quantum physics. Akin to an optical frequency comb, a beam can carry multiple various OAM components simultaneously. A series of discrete, equally spaced, and equally weighted OAM modes comprise an OAM comb. Inspired by the spatially extended laser lattice, we demonstrate both theoretically and experimentally an approach to producing OAM combs through azimuthal binary phases. Our study shows that transition points in the azimuth determine the OAM distributions of diffracted beams. Multiple azimuthal transition points lead to a wide OAM spectrum. Moreover, an OAM comb with any mode spacing is achievable through reasonably setting the position and number of azimuthal transition points. The experimental results fit well with theory. This work presents a simple approach that opens new prospects for OAM spectrum manipulation and paves the way for many applications including OAM-based high-security encryption and optical data transmission, and other advanced applications.
.About the Cover
The image on the cover depicts the generation of high-dimensional orbital angular momentum (OAM) comb by an azimuthal binary phase.