Integrated circuit (IC) manufacturing technology is the foundation of modern society. Accurate fabrication of chip design patterns faces challenges in pattern resolution, layer-to-layer overlay accuracy, and manufacturing yield. In particular, overlay error in IC manufacturing has been the critical factor for chip yield improvement. It is crucial for engineers to gain a comprehensive understanding of overlay errors, including their causes, measurement methods, feedback algorithms, and control elements. This review examines the technical challenges in chip manufacturing overlay alignment, particularly focusing on advanced process requirements for overlay error specifications. We address issues such as process variations leading to decreased overlay precision, reduced measurement accuracy, and increased difficulty in matching error control. The paper systematically analyzes methods and algorithms for improving overlay accuracy and control quality. These include measurement techniques, compensation models, mark selection, artificial intelligence integration, and self-aligned processes. By examining the relationship between process variations and chip overlay errors, this review provides valuable references for China's IC equipment and process development, aiming to enhance chip manufacturing yield through multi-factor collaborative development.

Polarization singularities are constructed by superposing orthogonal polarization circular Airyprime beams carrying different topological charges, generating a variety of initial polarization singularity structures such as quasi-lemon, V-point, and quasi-high-order polarization singularities. Numerical simulations are employed to investigate the evolution of these structures during free-space propagation. The results show that, for elliptical polarization fields, although the topological charges of the polarization structures remain invariant, their polarization states undergo significant evolution, characterized by changes in ellipticity, polarization handedness, and the transformation from quasi-singularities to standard polarization singularities. In contrast, for vector optical fields, the V-point structure remains topologically stable during propagation, while low-order vector fields can also evolve into stable polarization singularities. This work enriches the understanding of the formation and evolution mechanisms of polarization singularities and provides a theoretical basis for the precise control of polarization fields, with potential applications in optical micromanipulation, polarization encoding, and structured light field modulation.

Monocular depth estimation is a critical task that infers scene depth information from a single image. It is widely applied in fields such as autonomous driving, medical imaging, and defense. Deep learning methods have significantly enhanced the representational capacity and prediction accuracy of these models, particularly excelling at handling complex scenes, multi-scale features, and dynamic objects, which are challenging for traditional methods. This paper systematically reviews monocular depth estimation methods based on deep learning, beginning with an introduction to the fundamental technical process of monocular depth estimation. Then, based on the type of supervision, deep learning methods for monocular depth estimation are categorized into three groups: supervised learning methods are reviewed in terms of network structure, auxiliary information, loss functions, and depth discretization; unsupervised methods are summarized based on cues such as image pairs, masking, visual odometry, auxiliary information, and generative adversarial networks; semi-supervised methods are explored with respect to image pairs, semantic information, and generative adversarial networks. Subsequently, the paper outlines the main monocular depth estimation datasets and commonly used evaluation metrics, and lists the quantitative evaluation results of several methods on these datasets. Finally, the paper discusses application examples of monocular depth estimation based on deep learning, and highlights the major challenges and potential development directions for the future.

We propose a stereoscopic object super-resolution neural network (SSRNet) model based on pose estimation. The model uses a multi-task parallel network architecture to perform structural analysis and morphological feature extraction on low-resolution target images. The structural analysis estimates the target's pose and segments its components, while the morphological features are used to reconstruct the super-resolution image. This approach addresses the issue of varying target appearances under different poses, and fully utilizes a priori knowledge of target morphology, and achieves high-precision super-resolution image reconstruction as well as the identification and localization of key components and key points. This provides a new and effective technical approach for high-precision detection, recognition, and key component inspection of long-range stereoscopic small targets.

This study proposes an adaptive filtering non-local mean image (NLM) denoising algorithm based on the gradient of two-dimensional discrete functions and the optical intensity loss along the optical fiber, and applies it to the distributed optical fiber sensing system based on Brillouin scattering. On the basis of non-local mean filtering, in order to improve the signal-to-noise ratio of the sensing data of the Brillouin optical time-domain reflectometer (BOTDR) and solve the problem that the large difference in the noise standard deviation between the sensing rear end and the front end caused by the loss along the optical fiber affects the measurement accuracy and spatial resolution, this algorithm establishes relationships between the filtering coefficient and three parameters [the vertical gradient operator of Brillouin gain spectrum (BGS), the sensing distance, and the loss along the optical fiber] through a mathematical model. The denoising intensity is concentrated near the peak of the BGS and increases with the distance, so as to design an image noise reduction algorithm suitable for the distributed optical fiber sensing system. The research results show that compared with the original NLM algorithm, the proposed algorithm achieves an 8.56 dB improvement in the signal-to-noise ratio, and the standard deviation of the Brillouin frequency shift (BFS) is reduced by 0.91 MHz. In the face of the common multi-location and remote temperature change situations in engineering applications, the influence of the denoising of the proposed algorithm on the temperature accuracy is reduced by 15.29 percentage points, and the spatial resolution loss is controlled within 10%. In the field of distributed optical fiber sensing applications, the proposed algorithm improves the performance and cost-effectiveness of Brillouin detection instruments, and enhances its engineering practicability.

The traditional optical differentiators utilize interference, diffraction, or focusing properties to differentiate light waves, often requiring a complex optical setup for practical implementation. We propose a novel theoretical framework for an optical differentiator that achieves edge enhancement by precisely manipulating the orbital angular momentum (OAM) of pump light. By modulating the OAM of the pump light, higher-order differentiation of parametrically down-converted light waves is realized. Furthermore, leveraging the correlation properties of generated photon pairs by spontaneous parametric down-conversion (SPDC), we demonstrate quantum correlation imaging with edge enhancement. Based on quantum entanglement-based correlated imaging which enables non-local ghost imaging modality, we incorporate a high-order differentiator into the correlated imaging system to efficiently extract key features of images, thereby achieving non-local quantum image processing. The combination of these two techniques establishes a flexible and high-performance quantum optical differentiation system, providing a new technical pathway for optical computing and holding potential applications in future quantum image processing.

Silicon-based optoelectronic integration technology fully utilizes the advantages of low power consumption, fast response, and low signal crosstalk of optical interconnects. This technology is one of the significant solutions to break through the bottleneck of current development in electrical interconnect integrated circuits. Among them, silicon-based optical waveguide amplifiers containing rare-earth ions play a key role in achieving high-power and low-loss on-chip optical interconnects as the core device. This article reviews the research progress of silicon-based optical waveguide amplifiers and discusses three main research directions: silicon-based hybrid integrated III-V semiconductor optical waveguide amplifiers, silicon-based nonlinear Raman optical amplifiers, and silicon-based rare-earth ion-containing optical waveguide amplifiers. The research focuses specifically on silicon-based rare-earth ion optical waveguide amplifiers, exploring in detail the development history of their gain materials, waveguide structures, and integration platforms. The advantages and challenges of this device in providing optical gain and promoting single-chip silicon optical integration are also elaborated. We point out that rare-earth ion waveguide amplifiers are the key to promoting the practical application of silicon-based optoelectronic integration, and their development will provide important technical support for high-tech fields such as optical computing and data centers in the post-Moore's era.

Based on the fundamental principles of nonlinear optics, the properties of terahertz (THz) radiation generated through the optical rectification of shaped ultrashort laser pulses interacting with lithium tantalate (LiTaO3) crystal are systematically studied. The influence of laser pulse shaping parameters on the amplitude and spectral distribution of THz radiation is analyzed through numerical calculations. A 4F pulse shaping technique is employed to change and control the optical rectification process in LiTaO3 crystal, thereby changing the spectrum and amplitude of the generated THz pulses. The results reveal that due to the phonon absorption of the crystal at around 4.4 THz, the THz radiation produced by optical rectification can be divided into high-frequency and low-frequency bands. With shaped ultrashort laser pulses, both the central frequencies and amplitudes of the high-frequency and low-frequency components of THz pulses can be modified, enabling the generation of spectrally tunable THz pulses. This study provides a theoretical foundation for developing tunable broadband THz pulse radiation based on the interaction between shaped ultrashort laser pulses and crystals. Additionally, it offers a practical solution for this development.