In this paper, an optoelectronic oscillator (OEO) with a voltage-controlled phase shifter-based phase-locked loop (VCP-PLL) has been designed, analyzed and experimentally investigated. The experimental results show that the free-running OEO produces a frequency drift of 36 kHz within 60 min, whereas that of the designed VCP-PLL OEO is within 400 Hz. At frequency offsets of 1 kHz, and 10 kHz, the phase noise of the free-running and VCP-PLL OEOs are −76.1 dBc/Hz and −79.5 dBc/Hz, −98.4 dBc/Hz and −100.5 dBc/Hz, respectively. Additionally, in comparison with the OEO with a piezoelectric ceramic transducer-based PLL (PZT-PLL), the frequency of the VCP-PLL OEO can be stable for a longer time. However, both the VCP-PLL and PZT-PLL OEOs have lower phase noise compared to the free-running OEO, while the VCP-PLL OEO exhibits the lowest phase noise at low frequency offsets.

In this paper, the effect of a single-opening split-ring resonator (SRR) structure on the performance of ultra-wideband (UWB) microstrip patch antenna (MPA) is studied. Here a UWB MPA based on partial grounding is proposed, and the SRR structure is utilized to optimize other characteristics of the low-profile microstrip antenna while ensuring the bandwidth and gain remain unchanged. The proposed metamaterial structure, consisting of a 2×2 array of SRR, is loaded on the reverse side of the substrate. The volume of the proposed antenna is 61.6 mm×67.6 mm×1.5 mm. According to simulation and experimental results, the proposed microstrip antenna loaded with 2×2 SRRs performs better than that without SRR structure. It produces an operating band of 1.62–8 GHz while achieving circular polarization characteristics with an axial ratio (AR) bandwidth of 3.92–4.57 GHz. In addition, the performance of the proposed antenna is superior to relevant works of literature. The antenna can be widely utilized in wireless communication systems.

An underwater wireless optical communications (UWOC) system using different non-return to zero (NRZ) modulation schemes is proposed. Dual X- and Y-polarized Laguerre–Gaussian (LG) modes {LG[0,3], LG[1,3]} are realized at 10 Gbit/s data rate. It is found that the system using LG[1,3] offers 4.2 m range by using non-chirped NRZ modulation at −21.7 dBm receiver sensitivity and it can sustain 4.5 m distance with 1.7 mrad divergence, 0.1 dB transmitter and 0.5 dB receiver loss. Compared to existing UWOC systems, this design offers optimum performance with 8.2–9.5 µm photodiode (PD) width, −11.47 dB gain and −22.46 dBm signal and −79.10 dBm noise at receiver under clear and costal ocean scenarios.

In the field of laser speckle measurement, background interference has always been a major factor affecting the accuracy of measurements. The traditional Fourier analysis method can eliminate the background noise with certain periodicity, but the processing effect is not good for the complex background, and the operation of the signal is easy to lead to new errors. This study explores the application of a speckle differential measurement method in laser projection images through experiments with one-dimensional and two-dimensional sinusoidal signals, as well as step signals, combined with data simulation and actual measurement verification. The results demonstrate that this method effectively mitigates the interference caused by screen structure, brightness inhomogeneity, and pixel structure on measurement outcomes. Furthermore, it significantly enhances the accuracy of speckle contrast measurement. This research provides a simple and reliable new approach for the accurate measurement of speckle contrast.

Daytime optical observation has always been an important application in spatial situational awareness. Research has found that interference detection techniques can effectively suppress the influence of the skylight background and improve the detection capability of dim targets. However, it is intuitively believed that the interference detection system only has a small field of view (FOV), which is inconsistent with the experimental results (at least an FOV of 20°). This paper will study the reason why the shearing interference detection system can achieve imaging with a large FOV, provide the theoretical analysis conclusions and the simulation results, and then point out that the FOV of the interference detection system can be greater than 20°. A 97° FOV can form image plane interference.

Employing deep learning for pavement crack detection can significantly enhance accuracy, and federated learning can help to overcome the challenges of data silos and data security. This paper proposes a lightweight method to detect cracks. We use lightweight encoder modules to extract multi-scale features, and further feature fusion and modelling by selective fusion blocks and irregular-aware blocks. Moreover, this method combines federated learning with crack detection, effectively resolving the tension between data privacy and data sharing in decentralized devices. Experiments were conducted to compare our proposed method with other crack detection methods on two publicly available datasets, including the original method without federated learning and five state-of-the-art methods combined with the same federated learning framework. The experiment proves that our model obtains comparable F1 score with minimal parameters and computational effort. For example, compared to the original method, the number of parameters and computation are reduced by 67.1% and 73.6%, respectively.

In the process of crop growth, real-time detection and treatment of leaf diseases is extremely important. To improve detection accuracy and reduce the waste of resources, a you only look once version 8 nano (YOLOv8n) crop leaf disease detection model is proposed. First, we use the wise intersection over union (WIoU) loss function to intelligently focus on the model anchor boxes, accelerating model convergence. Next, to accommodate the size of the target being inspected, the detection heads are replaced to improve model accuracy. Then, we add the multi-head self-attention (MHSA) mechanism to the backbone network, enriching the feature extraction capability of the model for images. At last, we introduce deformable convolutional networks version 2 (DCNv2) in cross stage partial feature fusion (C2f) to improve the model’s ability to accurately localize input image targets. The experimental results show that the YOLOv8n_leaf disease model accuracy is improved by 1.3%, recall rate is improved by 0.4%, mean average precision at IoU threshold 0.5 (mAP0.5) reaches 83.7%, and mean average precision at IoU threshold 0.95 (mAP0.95) reaches 69.7%. The improved YOLOv8 model can provide technical support for crop leaf disease detection.

Small size, complicated background, and inaccurate target positioning are characteristics of remote sensing targets in the field of target detection. These factors might impair the detection network’s effectiveness and result in misidentification and omission. A module called LSKB_ECA is designed based on the you only look once version 8 (YOLOv8) backbone detection network. It combines the large selective kernel block (LSKB) with the efficient channel attention (ECA) mechanism to dynamically adjust the spatial sensing field and the localization of the key points, enhancing the model’s attention to the target’s features. A module called C2f_C-DCN is designed to replace the cross stage partial feature fusion (C2f) module in YOLOv8 in order to improve the model’s detection ability in complex scenes. The experimental results demonstrate that the enhanced YOLOv8 algorithm boosts the mean average accuracy by 2.7% and 2.1% on the dataset for object detection in aerial images (DOTA) and detection in optical remote sensing images (DIOR) datasets, respectively, signifying a notable enhancement in performance.

To enhance small object detection in complex road senses, an improved you only look once version 8 (YOLOv8) algorithm is introduced. Firstly, the quadruple down-sampling branch is added to improve the learning ability of the network for small object features. Secondly, a spatial pyramid pooling fast bi-level routing spatial attention (SPPF-BRSA) module is designed to remove irrelevant regions in a query adaptive manner, which effectively reduces the interference of complex background on detection performance. In addition, the C2fDynamic module is used in the neck of YOLOv8 to strengthen the feature expression ability of the model by dynamically selecting the convolution kernel. Finally, the wise intersection over union (Wise-IoU) v3 loss function is used to obtain more accurate detection results by adjusting the gradient gain distribution. The experimental results show that on Huawei SODA10M dataset, the improved algorithm improves precision (P), recall (R), F1 score and mean average precision at IoU threshold of 0.50 (mAP50) by 4.6%, 4.1%, 4.3% and 5.4%, respectively compared with the original algorithm.

The photon detection capability and nonlinear response characteristic of quanta image sensors make them an optimal choice for high dynamic range (HDR) imaging. To suppress ghosting artifacts in dynamic scenes and obtain high-quality HDR reconstructed images using quanta image sensors, an HDR image reconstruction method based on multi-weight factor exposure bracketing is proposed in this paper. Initially, adaptive aligning and merging of bit-planes from the same exposure is performed to suppress noise and blur in the low dynamic range (LDR) reconstructed frame. Subsequently, the exposure-referred signal-to-noise ratio (SNR) and gradient information of the frames are computed, which facilitates the reconstruction of an HDR image using multi-weight factor exposure bracketing. Results indicate that the proposed method has better reconstruction quality and is particularly effective for HDR imaging in non-ideal dynamic scenes. Compared to the single-weight exposure bracketing method and threshold optimization method, the proposed method has the lowest log-scale mean squared errors (LMSE) in reconstructed images and the perceptually uniform peak SNR is improved by 9.7%, while the perceptually uniform structural similarity index (PU-SSIM) is also improved by 3.5%, respectively.