Taoran Yue , Xiaojin Lu , Jiaxi Cai , Yuanping Chen , Shibing Chu
{"title":"YOLO-MST: Multiscale deep learning method for infrared small target detection based on super-resolution and YOLO","authors":"Taoran Yue , Xiaojin Lu , Jiaxi Cai , Yuanping Chen , Shibing Chu","doi":"10.1016/j.optlastec.2025.112835","DOIUrl":"10.1016/j.optlastec.2025.112835","url":null,"abstract":"<div><div>With the advancement of aerospace technology and the increasing demands of military applications, the development of low false-alarm and high-precision infrared small target detection algorithms has emerged as a key focus of research globally. However, the traditional model-driven method is not robust enough when dealing with features such as noise, target size, and contrast. The existing deep-learning methods have limited ability to extract and fuse key features, and it is difficult to achieve high-precision detection in complex backgrounds and when target features are not obvious. To solve these problems, this paper proposes a deep-learning infrared small target detection method that combines image super-resolution technology with multi-scale observation. First, the input infrared images are preprocessed with super-resolution and multiple data enhancements are performed. Secondly, based on the YOLOv5 model, we proposed a new deep-learning network named YOLO-MST. This network includes replacing the SPPF module with the self-designed MSFA module in the backbone, optimizing the neck, and finally adding a multi-scale dynamic detection head to the prediction head. By dynamically fusing features from different scales, the detection head can better adapt to complex scenes. The [email protected] detection rates of this method on three datasets IRIS, SIRST and SIRST+ reach 99.5%, 96.4% and 91.4% respectively, more effectively solving the problems of missed detection, false alarms, and low precision.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112835"},"PeriodicalIF":4.6,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143735132","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Wei Xu , Weihe Xu , Nathalie Bouet , Juan Zhou , Hanfei Yan , Xiaojing Huang , Zirui Gao , Ming Lu , Yong S. Chu , Evgeny Nazaretski
{"title":"Machine learning-aided automatic recognition and precise localization of marker layers within multilayer Laue lenses (MLLs) for high-resolution X-ray nanofocusing","authors":"Wei Xu , Weihe Xu , Nathalie Bouet , Juan Zhou , Hanfei Yan , Xiaojing Huang , Zirui Gao , Ming Lu , Yong S. Chu , Evgeny Nazaretski","doi":"10.1016/j.optlastec.2025.112847","DOIUrl":"10.1016/j.optlastec.2025.112847","url":null,"abstract":"<div><div>We present a new method for the automatic recognition and precise localization of marker layers within multilayer Laue lenses (MLLs) used for high-resolution X-ray nanofocusing. This approach integrates image processing techniques with machine learning algorithms, encompassing multiple stages including marker layer identification, coarse localization, differentiation, fine localization, and profile fitting. By directly obtaining marker layer profiles, this method eliminates errors induced by various factors such as manual measurements and image misalignment. It is robust and effective in the presence of various image defects. The proposed method enhances and streamlines the characterization of MLL optics, enabling accurate assessment of zone placement and detailed multilayer profile analysis. Consequently, it advances the fabrication of MLL optics and improves their application in high-resolution X-ray nanofocusing.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112847"},"PeriodicalIF":4.6,"publicationDate":"2025-03-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143735133","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Longchao Cao , Jingchang Li , Qi Zhou , Wang Cai , Binyan He , Yahui Zhang
{"title":"In-Situ monitoring of porosity based on static and dynamic molten pool features in laser powder bed fusion","authors":"Longchao Cao , Jingchang Li , Qi Zhou , Wang Cai , Binyan He , Yahui Zhang","doi":"10.1016/j.optlastec.2025.112872","DOIUrl":"10.1016/j.optlastec.2025.112872","url":null,"abstract":"<div><div>Porosity is one of the most serious defects for laser powder bed fusion (LPBF) which is a promising metal additive manufacturing (AM) technique. Reducing porosity is essential to improve the mechanical properties of parts in high-end applications. While existing research primarily focuses on static molten pool features, such as size and shape, these methods overlook critical local details, leading to imprecise porosity monitoring. To address this gap, this study proposes a novel attention-based method for in-situ monitoring of porosity (AMNet) from hybrid static and dynamic features (also referred to as motion features) extracted from the molten pool. Dynamic or motion features represent the molten pool by computing the distance from its edge to the centroid in the image. In the proposed AMNet, we adopt a self-attention mechanism to extract representative features from static and motion features separately. Furthermore, cross-attention is introduced to fuse two kinds of features in the latent space for porosity detection. To verify the effectiveness of AMNet, we collect a dataset consisting of molten pool and porosity data with three categories. The proposed AMNet achieves superior performance with an accuracy of 99.4 % on our dataset. Even 92 % accuracy is obtained with only 20 % of the training set by AMNet. The inference time of AMNet, averaging around 6.25 ms per sample has the potential to enhance closed-loop quality control systems by providing valuable real-time feedback. These results demonstrate the effectiveness and efficiency of the proposed method in porosity detection during LPBF.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112872"},"PeriodicalIF":4.6,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143724425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Fajing Li , Lin Yu , Lemin Shan , Yurui Wu , Yiran Wang , Yifan Wang , Shouping Nie , Jun Ma , Caojin Yuan , Xiangxing Xu
{"title":"Multi-focus beams with adjustable three-dimensional morphing spatial structure","authors":"Fajing Li , Lin Yu , Lemin Shan , Yurui Wu , Yiran Wang , Yifan Wang , Shouping Nie , Jun Ma , Caojin Yuan , Xiangxing Xu","doi":"10.1016/j.optlastec.2025.112859","DOIUrl":"10.1016/j.optlastec.2025.112859","url":null,"abstract":"<div><div>The beams with axial multiple focusing characteristic have been proven highly valuable in contemporary optics. We propose a new multifocal three-dimensional morphing (M3DM) beam featuring adjustable three-dimensional(3D) morphing spatial structure. By modulating the phase spectrum of a circular Airy beam via a chirped phase and a twisted phase, the circular Airy beam is transformed into a M3DM beam with four focal points. Moreover, the continuous three-dimensional enclosed dark space between two focal points in the middle presents the bottle beam characteristics. The 3D enclosed bottle space and even the focus points exhibit diverse structures by tuning the parameters of the modulated phase. From the energy flow analysis, it is observed that the gradient of energy flow varies with the beam structures. Employing the spatial multiplexing method, an array of M3DM beams is generated, capable of producing four times the number of focal points in 3D space. Furthermore, the dimension and focal position of each M3DM beam element can be independently adjusted with a high degree of freedom. Consequently, the proposed beam with its unique multifocal and self-rotational properties demonstrates promising applications in multiparticle trapping in 3D orientation, bioscience, optical communication, and multiplanar imaging.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112859"},"PeriodicalIF":4.6,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143724179","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhenhua Tang , Jun-Lin Fang , Yu-Xiang Wu , Guanhua Wu , Si Yang , Yan-Ping Jiang , Xin-Gui Tang , Xiu-Juan Jiang , Yi-Chun Zhou , Ju Gao
{"title":"Low-temperature solution-processed amorphous-Ga2O3 optoelectric synapses for neuromorphic computing","authors":"Zhenhua Tang , Jun-Lin Fang , Yu-Xiang Wu , Guanhua Wu , Si Yang , Yan-Ping Jiang , Xin-Gui Tang , Xiu-Juan Jiang , Yi-Chun Zhou , Ju Gao","doi":"10.1016/j.optlastec.2025.112837","DOIUrl":"10.1016/j.optlastec.2025.112837","url":null,"abstract":"<div><div>Optoelectronic synaptic devices present a promising approach to address the limitations of the von Neumann architecture. In this work, the amorphous gallium oxide (a-Ga2O3) optoelectronic synaptic devices were prepared by using a cost-effective sol–gel technique at a relatively low temperature of 400 °C. The a-Ga<sub>2</sub>O<sub>3</sub> devices exhibit an exceptionally wide bandgap and a stable, persistent photoconductive effect, allowing for the effective emulation of short- and long-term plasticity, paired-pulse potentiation, and pulse time-dependent plasticity akin to biological synapses. A digital image recognition method and a clothing image recognition method based on the LeNet-5 neural network model also were developed, achieving recognition rates of 97.8 % and 78 %, respectively. These findings are expected to contribute to the advancement of artificial synaptic devices, neural networks, and computing systems capable of optoelectronic operations.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112837"},"PeriodicalIF":4.6,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143714683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Lei Li , Bing Li , Xiang Wei , Zhenhua Gao , Zhengxi Lu , Hao Guo
{"title":"Multi-dimensional tool wear detection using Non-Subsampled Shearlet Transform and enhanced pulse coupled neural network in optical microscopy reconstruction","authors":"Lei Li , Bing Li , Xiang Wei , Zhenhua Gao , Zhengxi Lu , Hao Guo","doi":"10.1016/j.optlastec.2025.112849","DOIUrl":"10.1016/j.optlastec.2025.112849","url":null,"abstract":"<div><div>The milling tool plays a pivotal role in the fabrication of components within the aerospace and various other industrial sectors. Consequently, it is essential to perform swift, accurate, and comprehensive evaluations of tool wear throughout the manufacturing and processing stages. Traditional methods for assessing tool wear often suffer from limitations due to their dependence on singular evaluation criteria and a lack of detailed wear information. To overcome these challenges, this study introduces a multi-dimensional tool wear detection system utilizing optical microscopy vision to capture fully-focused images of the tool and reconstruct the shape of the wear region, which is designed to enable efficient, high-precision, and holistic evaluation of wear parameters. Furthermore, Non-Subsampled Shearlet Transform (NSST) and an enhanced pulse coupled neural network (PCNN)are used to extract 3D depth information, which facilitate the creation of a high-precision tool depth map by mapping high-frequency subbands to different depth levels while simultaneously obtaining the fully-focused image. Additionally, an inspection criterion is established that encompasses a multi-dimensional evaluation of wear metrics, including wear value, area, and volume. Compared to standard equipment, the error of wear value was found to be less than 0.005 mm, and the error rate of area, and volume was less than 2.5 %. Experimental results demonstrate that the proposed method offers more comprehensive assessment metrics for evaluating tool wear. It can be used to offer valuable feedback of tool state for machining processes.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112849"},"PeriodicalIF":4.6,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143724426","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Efficient interpolation-based and terrain-adaptive hierarchical filter for ultra-large-scale point cloud over complex landscapes","authors":"Chuanfa Chen , Lianzhong Xu , Jinda Hao , Yanyan Li , Dongxing Chen","doi":"10.1016/j.optlastec.2025.112860","DOIUrl":"10.1016/j.optlastec.2025.112860","url":null,"abstract":"<div><div>Complex landscapes, typically characterized by outliers, objects with diverse structures, vegetation on steep slopes, and terrain discontinuities, pose significant challenges to traditional filtering methods, especially when processing ultra-large-scale point clouds. To address these challenges, this paper proposes an efficient interpolation-based and terrain-adaptive hierarchical filtering method. Specifically, a hybrid algorithm combining a moving-window detector with robust surface fitting is developed to optimize the selection of initial ground seeds. Subsequently, a weighted finite-difference-based Thin Plate Spline (TPS) method is introduced to generate reference ground surfaces, thereby improving computational efficiency and mitigating the impact of misclassified object points. Finally, a terrain-adaptive filtering threshold incorporating different orders of terrain roughness is designed to accurately extract ground points near terrain discontinuities. To evaluate the proposed method, extensive experiments were conducted on the ISPRS benchmark samples and the ultra-large-scale OpenGF dataset. Results demonstrate that our method achieves an average Kappa coefficient of 92.3% across all 15 ISPRS samples, outperforming 24 state-of-the-art filtering methods published since 2010. On the OpenGF dataset, the proposed method surpasses five classical filters, reducing the average total error by 34.1%–77.4% and improving the average Kappa coefficient by 2.8%–21.7%. Overall, this framework provides a robust solution for filtering ultra-large-scale point clouds in complex landscapes.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112860"},"PeriodicalIF":4.6,"publicationDate":"2025-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143704424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Dimensionality reduction calibration of a robotic grinding head’s 3D coordinate system using a 1D laser sensor","authors":"Baichun Li, Zhi Li, Chang Ge, Chang Liu, Sujing Qin, Zhenpeng He","doi":"10.1016/j.optlastec.2025.112831","DOIUrl":"10.1016/j.optlastec.2025.112831","url":null,"abstract":"<div><div>The demand for highly accessible grinding in confined spaces, such as integral blisks, has driven research and applications in robotic grinding and polishing using a grinding head. Achieving high-precision and low-cost calibration of the grinding head is a primary challenge. Consequently, a novel dimensionality reduction calibration method is proposed. This method employs a 1D point laser displacement sensor to calibrate the 3D grinding head coordinate system. Firstly, a hand-eye calibration method for the sensor is proposed. The “fitting and one-step rotation method” determines the rotation axis and angle through the fitting, followed by one-step rotation to make the flange plane parallel to the laser emission plane, thereby completing the sensor orientation calibration. The “two-step rotation method” aligns the flange center with the measurement origin of the sensor to complete the sensor position calibration. Subsequently, the 3D grinding head coordinate system is calibrated with the pre-calibrated sensor as an intermediary. The “two-step translation method” makes the grinding head’s end face parallel to the laser emission plane, completing the grinding head’s orientation calibration. The “spatial three-point method” aligns the grinding head center with the sensor’s measurement origin, completing the grinding head’s position calibration. Finally, the experiments demonstrate that the precision of hand-eye calibration can be maintained within 0.178 mm and 0.6°, while the accuracy of the grinding head calibration is within 0.08 mm and 0.4°. The calibration method provides high precision, low cost, and reliable stability while supporting programming for rapid calibration, thereby facilitating the high-precision automation of robotic grinding and polishing.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112831"},"PeriodicalIF":4.6,"publicationDate":"2025-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143704425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Heng Wang , Qiang Cao , Tianhao Wu , Zhenzhong Wang , Yuting Hou , Sheng Peng , Du Wang
{"title":"Experimental investigation on crack evolution and separation strength in 4H-SiC slicing by sub-nanosecond laser","authors":"Heng Wang , Qiang Cao , Tianhao Wu , Zhenzhong Wang , Yuting Hou , Sheng Peng , Du Wang","doi":"10.1016/j.optlastec.2025.112826","DOIUrl":"10.1016/j.optlastec.2025.112826","url":null,"abstract":"<div><div>Compared with wire sawing technology, laser slicing technology has been considered a more advantageous technique for slicing SiC wafers with the advantages of small kerf loss and high efficiency. How to realize laser slicing of SiC with low separation strength and low damage remains a current challenge. In this paper, semi-insulating SiC slices with a separation strength of only 0.04 MPa and a damage layer thickness of only 5 μm were realized using a sub-nanosecond laser. The crack nucleation mechanism was delved into by analyzing the morphology and composition of the modified layer, and the influence of laser processing parameters on crack propagation was systematically studied. Tensile tests were performed on samples with varying crack states to uncover the relationship between crack state and separation strength. The results indicate that the damage layer comprises graphite and amorphous silicon, with secondary processing being the key factor in the nucleation of cracks on the cleavage surface (0 0 0 1) plane. Separation strength minimizes when the interval matches the crack-tip stress field boundary. And the feasibility of large-area processing was verified on 6-inch ingots.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112826"},"PeriodicalIF":4.6,"publicationDate":"2025-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143704423","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Optimization and arbitrary frequency stabilization of probe laser based on a transfer cavity in SERF Co-Magnetometer","authors":"Zhaoyang Cao , Xinxiu Zhou , Wenlei Zhao , Jingcheng Shang","doi":"10.1016/j.optlastec.2025.112875","DOIUrl":"10.1016/j.optlastec.2025.112875","url":null,"abstract":"<div><div>We demonstrate an arbitrary frequency stabilization technology that can ensure both the frequency stability and tunability of the probe laser in a spin-exchange relaxation-free (SERF) atomic co-magnetometer at the frequencies far from the atomic resonance line. In theory, the non-negligible coupling between the cavity length locking and the locked laser frequency is identified and a transfer function model for the frequency stabilization system is established. The modulation multiplexing cavity length locking method and dual-phase modulation sideband locking method are designed to overcome the limitations of traditional transfer cavity method, including noise coupling and insufficient frequency tunability. Experimental results show that at a frequency far from the atomic resonance line, the Allan deviation after the frequency locking reached 8.223 × 10<sup>−11</sup> for an averaging time of 10<sup>3</sup> s, which is lower by approximately two orders of magnitude than the frequency instability of free running laser. The research work is of great significance to improve the frequency stability, frequency tunability, cost-effectiveness, and compactness of the far-detuned probe laser in a SERF atomic co-magnetometer.</div></div>","PeriodicalId":19511,"journal":{"name":"Optics and Laser Technology","volume":"187 ","pages":"Article 112875"},"PeriodicalIF":4.6,"publicationDate":"2025-03-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143714682","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}