Optical Review最新文献

筛选
英文 中文
Wavelet-driven multi-scale feature extraction for underwater image restoration 基于小波驱动的水下图像多尺度特征提取
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-12-09 DOI: 10.1007/s10043-025-01018-2
Chao Pan, Yuxin Wu, Jing Zhang, Yan Wang, Xin Shu
{"title":"Wavelet-driven multi-scale feature extraction for underwater image restoration","authors":"Chao Pan,&nbsp;Yuxin Wu,&nbsp;Jing Zhang,&nbsp;Yan Wang,&nbsp;Xin Shu","doi":"10.1007/s10043-025-01018-2","DOIUrl":"10.1007/s10043-025-01018-2","url":null,"abstract":"<div><p>Underwater images suffer from low contrast, blur, and color deviation due to light absorption, scattering, and distortion. Traditional restoration methods, encompassing physical model-based compensations and basic filtering techniques, often struggle to adequately address these complex degradations. While deep learning approaches have demonstrated promising results on specific datasets, their ability to adapt to new conditions and operate reliably across varying illumination conditions and complex underwater scenarios remains challenging. To tackle these limitations effectively, this paper proposes a novel deep learning framework, termed WCM-Net, which incorporates wavelet transform, color compensation module, and dedicated multi-scale feature extraction block to decouple and enhance latent information at multiple frequency bands and spatial resolutions. ​​This integrated architecture efficiently learns comprehensive global-to-local feature representations for underwater image restoration. Experimental results on multiple public underwater datasets demonstrate that the proposed WCM-Net effectively enhances visual quality and significantly improves performance in downstream tasks such as key point detection and edge processing. Compared to existing state-of-the-art techniques, our approach exhibits superior robustness and adaptability, offering a valuable solution for underwater image processing applications.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"33 1","pages":"29 - 43"},"PeriodicalIF":0.9,"publicationDate":"2025-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145704096","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Germanium-based reconfigurable terahertz metasurface design for ultrafast dynamic frequency and amplitude modulation 基于锗的可重构太赫兹超快动态调频调幅超表面设计
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-12-08 DOI: 10.1007/s10043-025-01017-3
Qiangguo Zhou
{"title":"Germanium-based reconfigurable terahertz metasurface design for ultrafast dynamic frequency and amplitude modulation","authors":"Qiangguo Zhou","doi":"10.1007/s10043-025-01017-3","DOIUrl":"10.1007/s10043-025-01017-3","url":null,"abstract":"<div><p>The dynamic manipulation of terahertz (THz) waves exhibits tremendous application potential in high-speed communication, high-resolution imaging, and highly sensitive sensing. However, existing THz metasurfaces still face significant challenges in achieving ultrafast modulation speeds, high modulation depths, and broad tuning ranges, making it difficult to meet the demands for real-time tuning in complex application scenarios. To address these issues, this paper proposes a novel THz metasurface designed by embedding germanium (Ge) materials at specific locations to construct a reconfigurable metasurface. Under femtosecond laser modulation, the unit structure connected by Ge bridges transitions from meta-atomic to molecular model, enabling efficient dynamic manipulation of the frequency and amplitude of THz waves. Results indicate that at 0.523 THz, the amplitude modulation depth reaches 99%, with a tunable frequency range of 54.5%. Benefiting from the ultrashort relaxation time of photo-generated carriers in Ge, the complete recovery time of the resonant frequency is approximately 7 ps, demonstrating ultrafast switching capability and good reversibility. The proposed Ge-based THz metasurface modulator exhibits excellent performance in terms of frequency tuning range, modulation depth, and response speed, promising to provide new insights for the design of multifunctional THz metasurfaces, with broad application prospects in optical switching, wavefront control, and high-sensitivity sensing.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"33 1","pages":"10 - 21"},"PeriodicalIF":0.9,"publicationDate":"2025-12-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145697051","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Intensity diffraction tomography compressed imaging based on residual neural fields 基于残差神经场的强衍射层析压缩成像
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-12-03 DOI: 10.1007/s10043-025-01011-9
Cheng Zhang, Xu Ping Zhang, Li Ru Zhang, Wei Wei Ma, Qing Ye, Guo Hao Zhu, Chuan Shen, Hong Cheng, Quan Bing Zhang, Sui Wei
{"title":"Intensity diffraction tomography compressed imaging based on residual neural fields","authors":"Cheng Zhang,&nbsp;Xu Ping Zhang,&nbsp;Li Ru Zhang,&nbsp;Wei Wei Ma,&nbsp;Qing Ye,&nbsp;Guo Hao Zhu,&nbsp;Chuan Shen,&nbsp;Hong Cheng,&nbsp;Quan Bing Zhang,&nbsp;Sui Wei","doi":"10.1007/s10043-025-01011-9","DOIUrl":"10.1007/s10043-025-01011-9","url":null,"abstract":"<div><p>Intensity diffraction tomography (IDT) encompasses a class of optical microscopy techniques designed to reconstruct the three-dimensional refractive index (RI) distribution of a sample from a series of two-dimensional intensity-only measurements. However, reconstructing artifact-free RI maps remains a fundamental challenge in IDT owing to the loss of phase information and the missing-cone problem. Neural fields (NF), a deep learning framework increasingly adopted in computer vision and graphics, offer a promising approach by using a neural network to map 3D spatial coordinates to optical properties—such as color, opacity, and refractive index. we propose a method called Intensity Diffraction Tomographic Compressed Imaging based on Residual Neural Fields (IDTCI-RNF). Our approach integrates a compressed sampling mechanism into the IDT imaging model and employs a self-supervised deep learning framework enhanced with 3D Total Variation (3DTV) regularization to preserve edge details. The proposed method enables high-quality reconstruction of 3D refractive index distributions from low-dimensional discrete intensity measurements. Experimental results demonstrate that IDTCI-RNF produces high-contrast, artifact-free RI maps and outperforms existing methods in retaining structural edges and fine details.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 6","pages":"779 - 790"},"PeriodicalIF":0.9,"publicationDate":"2025-12-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145657929","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Optical single-channel security system based on Schur decomposition and biometric random phase encoding in fractional Fourier transform domains 基于分数阶傅里叶变换域舒尔分解和生物特征随机相位编码的光单通道安全系统
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-11-21 DOI: 10.1007/s10043-025-01015-5
Muhammad Rafiq Abuturab, Aparajita Anand, Faiz Ahmad, Nadim Asghar, Weqas Juraize Moniri, Shahzad Ahsan, Ratnesh Kumar Singh, Md. Nesar Ahmad, Amal Kumar
{"title":"Optical single-channel security system based on Schur decomposition and biometric random phase encoding in fractional Fourier transform domains","authors":"Muhammad Rafiq Abuturab,&nbsp;Aparajita Anand,&nbsp;Faiz Ahmad,&nbsp;Nadim Asghar,&nbsp;Weqas Juraize Moniri,&nbsp;Shahzad Ahsan,&nbsp;Ratnesh Kumar Singh,&nbsp;Md. Nesar Ahmad,&nbsp;Amal Kumar","doi":"10.1007/s10043-025-01015-5","DOIUrl":"10.1007/s10043-025-01015-5","url":null,"abstract":"<div>\u0000 \u0000 <p>A new optical single-channel security system based on Schur decomposition, biometric random phase encoding in fractional Fourier transform domains is proposed. In this technique, the grayscale iris image is decomposed by Schur decomposition. Unitary matrix<i> U</i> and upper triangular matrix <i>T</i> are encoded and modulated independently by a random phase mask to produce a biometric random phase mask (BRPM). An original color image is split into <i>R</i>-, <i>G-</i> and <i>B-</i> channels. Each channel is decomposed by Schur decomposition. The unitary matrix<i> U.</i> of each channel is combined to construct a single-channel image as an input image. The upper triangular matrix <i>T</i> of each channel is utilized as decryption key. A single-channel image is modulated with <i>U-</i> BRPM and then fractional Fourier transformed. The fractional Fourier spectrum is amplitude- and phase-truncated to obtain first encrypted image and first decryption key. The first encrypted image is modulated with <i>T-</i> BRPM and then fractional Fourier transformed. The fractional Fourier spectrum is amplitude- and phase-truncated to get final encrypted image and second decryption key. The proposed cryptosystem has three main advantages. First, the proposed system has four decryption keys to evade potential attacks. Second, <i>U-</i> BRPM and <i>T-</i> BRPM are utilized as unique encryption keys to resist unauthorized access. Finally, the fractional orders of the fractional Fourier transform provide sensitive encryption keys. The proposed method can be executed using a hybrid optoelectronic system. The viability, efficiency, and security of the proposed system can be shown by numerical simulation results.</p>\u0000 </div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 6","pages":"816 - 829"},"PeriodicalIF":0.9,"publicationDate":"2025-11-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145567280","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Advanced circular dichroism measurements with coherent circular polarization switching in dual-comb spectroscopy 双梳光谱中相干圆偏振开关的先进圆二色性测量
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-11-12 DOI: 10.1007/s10043-025-01010-w
Ruichen Zhu, Akifumi Asahara, Takashi Kato, Kaoru Minoshima
{"title":"Advanced circular dichroism measurements with coherent circular polarization switching in dual-comb spectroscopy","authors":"Ruichen Zhu,&nbsp;Akifumi Asahara,&nbsp;Takashi Kato,&nbsp;Kaoru Minoshima","doi":"10.1007/s10043-025-01010-w","DOIUrl":"10.1007/s10043-025-01010-w","url":null,"abstract":"<div><p>We propose a novel method for measuring circular dichroism (CD) using polarization-modulated dual-comb spectroscopy (PM-DCS). This method utilizes two tightly phase-locked optical frequency combs to realize rapid and precise polarization modulation, eliminating the need for mechanical polarization modulators, and achieves high-speed, high-precision control of the polarization states. The experimental results demonstrated successful polarization switching between left- and right-circularly polarized light, enabling accurate detection of the CD spectra. Moreover, this method supports customizable polarization switching, realizing applications such as simultaneous circular-linear dichroism measurements. Our findings offer a pathway for high-sensitivity, broadband, and dynamic polarization measurements, extending applications in the analysis of chiral materials, particularly in the fields of chemistry, biology, materials science, and device characterization, where the ability to perform real-time high-sensitivity measurements is essential.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 6","pages":"843 - 850"},"PeriodicalIF":0.9,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10043-025-01010-w.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145492485","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Modeling high-speed long-haul DWDM/UDWDM networks with FWM minimization 基于FWM最小化的高速长途DWDM/UDWDM网络建模
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-11-12 DOI: 10.1007/s10043-025-01014-6
S. S. Supriya, Joseph Zacharias
{"title":"Modeling high-speed long-haul DWDM/UDWDM networks with FWM minimization","authors":"S. S. Supriya,&nbsp;Joseph Zacharias","doi":"10.1007/s10043-025-01014-6","DOIUrl":"10.1007/s10043-025-01014-6","url":null,"abstract":"<div><p>Information transmission capacity in optical fiber networks can be rapidly escalated using Dense Wavelength Division Multiplexing (DWDM). However, non-linearities in fiber such as Four Wave Mixing (FWM) cause adverse effects in DWDM and ultra-DWDM networks. In this work, differential quadrature phase shift keying (DQPSK) modulation is employed to minimize the impact of FWM effect. The system performance can be improved by properly adjusting the switching and bias voltage of the dual-port Mach-Zehnder Modulator (MZM). Here, the performance of the proposed method is compared with other advanced modulation techniques such as Quadrature Phase Shift Keying (QPSK), Differential Binary Phase Shift Keying (DBPSK) and 16-Quadrature Amplitude Modulation (QAM). The results reveal optimal performance with maximum FWM minimization for the proposed method at a switching voltage of 3 V.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 6","pages":"802 - 815"},"PeriodicalIF":0.9,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145492484","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Proposed novel relative intensity LIBS approach for quantifying heavy metals in sunscreens, compared with AAS 提出了一种新的相对强度LIBS方法来定量防晒霜中的重金属,并与原子吸收光谱法进行了比较
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-11-12 DOI: 10.1007/s10043-025-01013-7
Rajaa Nader, Muayyed Jabar Zoory, Haidar J. Mohamad
{"title":"Proposed novel relative intensity LIBS approach for quantifying heavy metals in sunscreens, compared with AAS","authors":"Rajaa Nader,&nbsp;Muayyed Jabar Zoory,&nbsp;Haidar J. Mohamad","doi":"10.1007/s10043-025-01013-7","DOIUrl":"10.1007/s10043-025-01013-7","url":null,"abstract":"<div><p>Concerns about sun exposure have increased, leading to increased interest in the quality of sunscreen products, particularly regarding their content of heavy metals that may pose a health risk. This study aims to evaluate the concentrations of some heavy metals in sunscreens using two analytical techniques: atomic absorption spectroscopy (AAS) and laser induced breakdown spectroscopy (LIBS), and to compare the results of both methods. Two commercial samples representing two different quality products were analyzed: a low-quality sample (R10) and a high-quality sample (R11). Sunscreen samples were dried at 55 °C for 1 h (before LIBS analysis) to remove excess moisture and to provide a stable plasma, as the sunscreen matrices had high contents of water and oil. Two computational methods were used to estimate element concentrations The first is an electron density method to determine the concentrations of elements such as Zn, Ni, Pb, Hg, and Cd. The second is a new method known as the relative density method (LIBS-Qc), a simplified method that does not rely on electron temperature and has proven highly efficient in calculating concentrations. The results showed that the mercury concentration was higher in sample R10 (11.083) than in sample R11 (6.607). The new method (LIBS-Qc) also showed good agreement with the results of the AAS technique, indicating its potential as an effective and rapid alternative for the analysis of heavy elements in commercial products.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 6","pages":"791 - 801"},"PeriodicalIF":0.9,"publicationDate":"2025-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145492489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
An closed-cavity Mach–Zehnder interferometer based on multi-core fiber with two-end etching for relative humidity 一种基于多芯光纤两端刻蚀的闭式马赫-曾德尔干涉仪
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-10-23 DOI: 10.1007/s10043-025-01000-y
Junni Cheng, Xiao-yan Jiang, Yu-heng Wang
{"title":"An closed-cavity Mach–Zehnder interferometer based on multi-core fiber with two-end etching for relative humidity","authors":"Junni Cheng,&nbsp;Xiao-yan Jiang,&nbsp;Yu-heng Wang","doi":"10.1007/s10043-025-01000-y","DOIUrl":"10.1007/s10043-025-01000-y","url":null,"abstract":"<div><p>A Mach–Zehnder interferometer (MZI) based on the two-end etching of multi-core fiber (MCF) is proposed and experimentally demonstrated. The MZI consists of a seven-core fiber with two-end etching and is sandwiched in two single-mode fibers. To study the influence of different etching time on MZI, BOE solution is used to etch MCF. Due to the fact that BOE solution is a mixture of HF and NH<sub>4</sub>F, the etching rate of BOE on the MCF is slower, which is more conducive to observing the etching appearance of the MCF. The effect of etching on MZI spectrum is analyzed theoretically. Experimental results confirm and show that the humidity increases from 25%RH to 95%RH, the MZI energy monitoring point decreases, and its humidity response sensitivity is − 0.045 dB/%RH. Since the temperature also affects the sensor response, the temperature is also measured. The temperature increases from 35℃ to 85 ℃, and the MZI temperature response sensitivity is 0.0465 nm/℃. A quick and reliable time response has also been demonstrated and shows potential for future applications.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 5","pages":"674 - 683"},"PeriodicalIF":0.9,"publicationDate":"2025-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10043-025-01000-y.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145384799","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Multi-band design of ultrathin terahertz metamaterial filter based on square-loop arrays 基于方环阵列的超薄太赫兹超材料滤波器多波段设计
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-10-08 DOI: 10.1007/s10043-025-01008-4
Anqi Li, Zishan Yang, Zhitao Wang, Zhaoyang Chen, Jian Jing
{"title":"Multi-band design of ultrathin terahertz metamaterial filter based on square-loop arrays","authors":"Anqi Li,&nbsp;Zishan Yang,&nbsp;Zhitao Wang,&nbsp;Zhaoyang Chen,&nbsp;Jian Jing","doi":"10.1007/s10043-025-01008-4","DOIUrl":"10.1007/s10043-025-01008-4","url":null,"abstract":"<div><p>This paper presents several ultrathin multi-band terahertz bandpass filters (BPFs) based on metal-dielectric-metal (MDM) metasurfaces, achieving independently tunable multi-band transmission characteristics through parametric adjustment of the geometric parameters (side length <i>w</i>, gap spacing <span>(varDelta s)</span>, and loop count <i>N</i>) of the square-loop arrays. The designed tri-band filter is primarily analyzed for its excellent performance at 0.72/1.23/2.23 THz, including a roll-off rate &gt;30 dB/THz, average transmittance of 97% (peak insertion loss <span>(&lt; -0.10)</span> dB), and out-of-band rejection ratio &gt;25 dB. The multi-band spectral response is validated via finite integration technique (FIT) simulations in CST Microwave Studio, combined with surface electric field distribution analysis and effective medium theory (EMT) to elucidate the propagation mechanism of THz waves and the functional roles of each structural layer. A systematic investigation of geometric parameter dependencies on transmission performance is also conducted. The designed BPF exhibits potential applications in electromagnetic interference suppression, ultrasensitive biosensing, and dynamic THz signal multiplexing, owing to its multi-band transmission, polarization insensitivity, and compact footprint.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 5","pages":"745 - 761"},"PeriodicalIF":0.9,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145384798","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
Improving human pose recognition with YOLOv8-SLG using thermal images 利用热图像改进YOLOv8-SLG的人体姿态识别
IF 0.9 4区 物理与天体物理
Optical Review Pub Date : 2025-10-02 DOI: 10.1007/s10043-025-01009-3
Yuqian Zhang, Zhan Wang
{"title":"Improving human pose recognition with YOLOv8-SLG using thermal images","authors":"Yuqian Zhang,&nbsp;Zhan Wang","doi":"10.1007/s10043-025-01009-3","DOIUrl":"10.1007/s10043-025-01009-3","url":null,"abstract":"<div><p>Thermal imaging cameras operate effectively at night or in low-light conditions, enhancing the accuracy of human pose recognition in such environments. To address the issue of insufficient accuracy in traditional human pose recognition methods under low-light conditions, this study proposes an improved algorithm based on the YOLOv8 model, referred to as YOLOv8-SLG. Considering the peculiarity that thermal images reflect temperature information rather than optical features, traditional convolutional networks may produce unnecessary redundancy in feature extraction. To address this issue, first, we improve the detection accuracy and preserve the network structure by replacing the original convolution with SCConv building blocks in the backbone network of YOLOv8n in order to reduce spatial and channel redundancy between features in the convolutional neural network. Second, we enhance local feature detection by integrating the LSKA attention mechanism into the neck network, reducing computational complexity and memory requirements while maintaining accuracy. Finally, we enhance the multi-scale processing capability and reduce the number of parameters per detection head through shared GroupNorm convolution to improve target localisation and classification performance. Experimental results show that these enhancements significantly improve the model’s performance for human pose recognition tasks in complex contexts. Compared to the original YOLOv8n model, the proposed algorithm improves the precision, recall, mAP50, and mAP50-95 metrics by 1.33%, 1.79%, 1.86%, and 2.01% to 97.7%, 94.6%, 96.7%, and 75.1%, respectively. In addition, YOLOv8-SLG reduced model parameter calculations by 8.14%. It can detect human poses in thermal images in real time accurately, and comparison with other mainstream human pose detection algorithms confirms the effectiveness and superiority of the method.</p></div>","PeriodicalId":722,"journal":{"name":"Optical Review","volume":"32 5","pages":"762 - 778"},"PeriodicalIF":0.9,"publicationDate":"2025-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145384797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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}
引用次数: 0
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
相关产品
×
本文献相关产品
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信
小红书